Dingo 4.1 RC Bugs

[liberomureddu]

1) the script is not in /usr/bin, but in /usr/sbin. a difference.

Ooops, thanks for correcting this, sorry.

[liberomureddu]

Sometimes the volume and mic are not set. Use ALSAMIXER to fix. You can find it at the start menu>multimedia>alsamixer

Yes it works, it was, in fact, a wrong alsamixer setting.
thanks.
libero

[Béèm]

I am (on a virtual PC, qemu) experimenting with puppy in full install.
I saw that when calling the bootmanager configure bootup and clicking on selection of sfs files I get a message saying: no sfs file found in /mnt/home.

Well yes that's logical. The way of using/installing a sfs in a full install is different.

The bootmanager script should detect it's a full install and give instructions accordingly.

[Scoticus]

I downloaded the RC last night and was unable to boot the system after the initial boot.

As usual Puppy will not properly display as I have Nvidia graphics and consequently I have to download the drivers.

In normal circumstances I use the setup menu and set the firewall on followed with the network wizard to access the web. Having got that far I then download the Nvidia driver and reboot the system saving the information as a multisession disk.

On rebooting this time the programme went through the usual cycle only to display an error message. I was asked to use xorgwizard which was a fat lot of good as when I reset the display size it reverted back to the error.

This is the first time that I have had problems with the display since I was informed of the need to load Nvidia drivers.

Can anyone offer a solution to this problem.

[tempestuous]

This is the first time that I have had problems with the display since I was informed of the need to load Nvidia drivers.

Which nVIDIA driver package are you referring to?
Puppy4.1rc has an updated 2.6.25.16 kernel, and I'm not aware of anyone having compiled a compatible nVIDIA graphics driver package for this kernel yet.

[BarryK]

With Barry's coaching, I have a fix for the icon problem I reported. It is consistent with Barry's hotplug strategy.

Please test it to see how it affects those having icons get overlaid.

Here is the difference file contents for clean_desk_icons:

Code: Select all

25c25,39
<  for ONEDRV in `ls -1 /sys/block | grep -vE 'loop|ram' | tr '\n' ' '`
---
>  #v410 Delete drive_ directories for removed drives (hd.*, sd.*, but sr.)
>  DRVS_PRESENT="`ls -1 /sys/block | grep -vE 'loop|ram'`"
>  DIR_DRVS="`ls -1 /root/.pup_event | sed s/drive_// | grep -E "^[hs]d[a-z]$|^sr[0-9]*$"`"
>  for ONEDRV in $DIR_DRVS;do
>   if [ "`echo "$DRVS_PRESENT" | grep -e "$ONEDRV"`" = "" ];then
>    [ "`echo "$ONEDRV" | grep -e "^sr[0-9]*"`" != "" ] && rm -rf /root/.pup_event/drive_$ONEDRV || rm -rf /root/.pup_event/drive_$ONEDRV*
>   fi
>  done
>  
>  #Postpone icons for optical drives
>  PROBEDISK="`probedisk2`"
>  DRVS_OPTICAL="`echo "$PROBEDISK" | grep '|optical|' | cut -f 1 -d '|' | cut -f 3 -d '/'`"
>  #v410 end
> 
>  for ONEDRV in `echo "$DRVS_PRESENT" | tr '\n' ' '` #v410
27,28d40
<   #[ "$ONEDRV" = "sr0" ] && continue
<   #[ "$ONEDRV" = "sr1" ] && continue
34,35c46
<    [ ! -d /root/.pup_event/drive_${ONEDRV} ] && continue #v408
<    echo "$OKDRV" >> /tmp/pup_event_ok_pin
---
>    [ -d /root/.pup_event/drive_${ONEDRV} -a "`echo "$DRVS_OPTICAL" | grep -e "$ONEDRV"`" = "" ] && echo "$OKDRV" >> /tmp/pup_event_ok_pin #v408 v410

Here is the difference file contents for pup_event_frontend_d:

Code: Select all

327a328
>  [ "$CDSTATUS2" != "0" ] && rm -rf /root/.pup_event/drive_${DRV_NAME} #v410 ensure no drive directory if drive empty

I have tested this on both the standard (2.6.25.16) and retro (2.6.21.7) kernel release candidates.
Richard

EDIT 9/30: Updated dotpet and diff listings to support cleanup for removable optical drives in standard kernel. (Did not look for drives "sr...".) Since I have no such drive, anyone with a USB CD device: please try a "mount then shutdown-unplug-reboot" test on the standard kernel -- it should already work for the retro kernel.
Richard

Richard,
This is my take on it, with ideas from you. Also I added improved collision-avoidance so that new icons cannot be created nearly on top of another desktop icon.

[davesurrey]

Hi Barry,
Excuse me butting in here but will these files do anything to help my problem with 4.1rc (full install, latest 2.6.25 kernel) which I reported in beginners section.

Now realise I should have done this here but at the time thought it might be my error so didn't want to raise as a bug.

Prob is when I try to rename desktop partition drive icons (eg sda1 to Windows) I get two sets of icons (Windows AND sda1.)

In adddition can't see any way to line up the icons in a row (trivial prob by comparison).

Cheers and very many thanks for all your wonderful work.
Dave

[davesurrey]

@Scoticus
@tempestuous

Don't kinow if this will help but I am running puppy 4.1rc (2.6.25.16 kernel) and have had no problems running with my nVidia card. Same as with Dingo 4.0. No tweaks or installs ever needed.
Card is an nvidia GeForce 4 (MX440), a bit old I know but works for me.
Cheers
Dave

[boscobearbank]

In adddition can't see any way to line up the icons in a row (trivial prob by comparison).

I've taken to editing /root/Choices/ROX-Filer/PuppyPin manually. This obviously is not a good solution; just a crude but effective work-around.

[davesurrey]

Frankly that's such a retro step back from 4.0 where the icons could be positioned in various places on the screen, they lined up automatically and I could change their name without having double icons for every partition. Bring back HotPup until this new Event Manager is working well is my vote.

[BarryK]

In adddition can't see any way to line up the icons in a row (trivial prob by comparison).

I've taken to editing /root/Choices/ROX-Filer/PuppyPin manually. This obviously is not a good solution; just a crude but effective work-around.

Re-alinging the icons, and putting different names on them is something that can be added to the Event Manager GUI. I intended this for the next version, probably for Puppy 4.2. For 4.1, an incredible amount has happened "under the hood", and these kind of extra features are fairly trivial "bells and whistles" that have had a low priority, until we have got the underlying mechanisms working perfectly.

4.1 is basically "frozen" now, but requests will be considered for 4.2, in fact no need to wait for that, I can release an upgrade PET anytime.

[firak]

I test on eee 701 this new 4.1 . very ince ! i think i will muve to that.
but, becouse the eee had the small screen i very usefool to use the oauto hide jwm opcion.
Actually in the options it there, but does not really works on my eee . maaybe thereis a bug in the jwm set up .

[davesurrey]

Barry,
I fully understand why things must get prioritized and designs frozen.

Just wanted to see if I was doing anything wrong or whether it was possible to re-align and change names with the existing code.

Okay will wait but in the meantime if anyone else feels like making a pet to do this I sure would be grateful. Afraid it's above my capabilities.
Cheers
Dave

[tempestuous]

Don't know if this will help but I am running puppy 4.1rc (2.6.25.16 kernel) and have had no problems running with my nVidia card.

Yes, me too. This is the simple solution where we use the opensource Xorg "nv" driver which is standard in Puppy.
But there are some troublesome nVIDIA devices which fail to work with this driver, and require the more sophisticated 3D proprietary driver from nVIDIA. Clearly Scoticus has one of these difficult models.
Since there is a kernel module associated with the proprietary graphics driver, it's kernel-version sensitive.

[MU]

nvidia drivers and kernelheaders now available from gray:

http://www.murga-linux.com/puppy/viewtopic.php?t=34001
http://www.murga-linux.com/puppy/viewtopic.php?t=34003
Mark

[Flash]

This seems like a small thing to me but I want to mention it in case.

I have a "universal" flash card reader which I leave plugged into one USB port, and an 8 GB flash memory stick is kept plugged into the other USB port. Usually the "universal" reader has no flash memory plugged into it. I only use it to look at the pictures from my digital camera. When Puppy 4.1 RC boots, it puts an icon on the desktop for the 8 GB flash stick, as sde1, and one for the empty reader - as a 4 GB NTFS drive!

[rerwin]

With no disks in any of my CD/DVD drives, but with a USB flash drive plugged in, pmount shows the flash drive under its "optical" tab instead of "usbdrv". This only occurs when pmount is started from the "mount" icon or the menu; the drive icons force the usbdrv tab to be first.

Apparently, if there are only 3 device types to be displayed (plus the empty optical drives), the 4 tab labels have usbdrv last (in the tab-label string), but only the first three are used.

While this is beyond my ability to fix in short order, maybe Barry "just knows" what to fix. It seems related to the logic that skips inclusion of drives with no detectable file system (or perhaps other reasons, too), as is the case for empty optical drives. Those drives are unique in that, although the hardware is present, it is not treated as present until a disk is inserted. The tab labels are generated from the list of present hardware, but no consideration is given to the possibility that there may be no partitions that qualify for inclusion under the "optical" tab.

One simple solution might be to ensure that "optical" is always last in the tab list (unless forced to be first), so it wouldn't confuse the viewer of the pmount display.
Richard

Unrelated info: I have updated my hotplug icon fix to "-3", which includes Barry's collision avoidance fix, at:
http://www.murga-linux.com/puppy/viewto ... 793#235793

[PaulBx1]

I took my Puppy 2.16 pupsave, deleted all installed packages (that petget knew about anyway) and tried to upgrade it to 409. This did not go too well.

1) Got a message, "You will have to uninstall the Puppy 2 packages..." or gave me the option of installing all the dependencies like gtk1 and tcl/tk, which I did not want to do. I believe I have some dotpups that the package manager is unaware of even though the boot process detected something was wrong. It is hard to know where to start, to find packages I installed many months ago and forgot about. If the boot process has any idea what the packages are, (seems doubtful) it should say. Anyway I have some detective work on my hands. I suppose I can just dig through /initrd/pup_rw, looking for stuff.

2) It claimed there would be a versioncleanup directory under /tmp, as usual, but none was there.

3) There were a variety of error messages in the logs, "file not found", that sort of thing. To be expected I suppose.

I hope this upgrade is possible. Would hate to start from scratch...

[linuxcbon]

k2.6.25.16 live CD pfix=RAM

freememapplet shows only 1Gb RAM instead of 2.

dmesg gives
ohci_hcd: Unknown symbol ssb_device_disable
ohci_hcd: Unknown symbol ssb_admatch_base
ohci_hcd: Unknown symbol ssb_device_enable
ohci_hcd: Unknown symbol ssb_driver_unregister
ohci_hcd: Unknown symbol __ssb_driver_register
ohci_hcd: Unknown symbol ssb_admatch_size
--Seems to do with b43 and ssb modules
see https://bugs.launchpad.net/ubuntu/+sour ... bug/182716
seems to be solved with kernel 2.6.27-rc5-r1

/usr/etc what is it ?

/etc/gadmin-rsync is empty

Can you add firmware file dvb-usb-dib0700-1.10.fw for DVBT
Only 34kb
Can be found at
http://www.wi-bw.tfh-wildau.de/~pboettc ... 00-1.10.fw
Very useful for many usb sticks.
To be in /lib/firmware/

[PaulBx1]

When you click on a link to a postscript file, Seamonkey displays the usual dialog about not knowing what to do with it, and suggests the "default application (ghostscript)". If you choose that you get it displayed in epdfview instead.