Porteus-v1.0-rc3-x86_64 "Unplanned release"

[fanthom]

After short discussion with brokenman we have decided to release rc3 for both arch for one week testing before going with 1.0 milestone.
This is the last chance for reporting bugs so every feedback is welcome

Link to the repo:
http://ponce.cc/porteus/x86_64/testing/ ... -v1.0_rc3/

Short changelog:
- kernel 2.6.38.8
- KDE-4.6.4
- added super hot feature: 'Porteus Package Manager' by brokenman!
- linuxrc supports booting from 'UUID:' and 'LABEL:' which can be used instead of /dev/sdXy for all cheatcodes which uses full paths. more info:
http://porteus.org/forum/viewtopic.php?f=53&t=569
thanks to y1b for inspiration!
- removed creation of .dat file by linuxrc. please use 'Porteus .dat manager' instead
- 'noeject' cheatcode has one additional feature when used together with copy2ram: prevents unmounting of booting partition
(please read /boot/chatcodes.txt for more detailed info)
- fixed 'load=' and 'noload=' cheatcodes which were conflicting in some cases (thanks to crashman for bug report)
- Ahau set all cheatcodes in alphabetical order (francois suggestion) - thanks for that.
- /boot/tools/TweakUAC.exe replaced with MyUAC by Blaze (thanks!)
- removed 'quiet splash=' cheatcodes from liloinst.sh as we dont use splash
- added user 'guest' to following groups: power and dialout
- kernel config: RAID is compiled as (M) . initrd doesn't support it anyway.
- added more info to the /001/opt/porteus-scripts/porteus_encrypter script. should be more user friendly now.
- 'Porteus module manager': fixed deactivation of modules
- removed all firmware which had no license attached, added atheros fw
- lzm2xzm: fixed convertions of folders with "lzm" in name
- porteus-usb.sh: calls lin_start_here.sh and not bootinst.sh at the end
- psinfo: added '-k' switch to lspci which shows kernel drivers used for pci hardware
- added licenses to /usr/doc for all non-free packages (ie: firmware, xvba-video)
- 'Porteus Module Tools' got various fixes: merged tgz and txz options, additional info is displayed after creating xzm from changes
- /002/usr/share/applications/cups.desktop: changed path to http://localhost:631/admin (thanks to wread)
- 'Porteus .dat manager': added info about "changes=" cheatcode and .dat containers
- /002/xf86-video-nouveau-git is downgraded to previous version as per Blaze's bug report

Be aware that 64bit repo for 'Porteus Packages Manager' is not populated with packages yet. i have got 64bit version of PPM 30 mins ago. currently we have 32bit ones there so you can use them for testing. Give me few days and i'll start replacing 32bit ones with 64bit equivalents.

Happy testing!

The Porteus Team

[kalo86]

Hi fanthom,
with the new release RC3 I had 3 freezes on my laptop. They happened when I opened the file manager Dolphin, exactly when I was scrolling into the pendrive folder.
The pendrive's root (USB) contains the /boot and /porteus folders. It happened 3 times. The mouse still moves, but no clicks on objects or buttons. The only key which works is the power button.
I hope my description may help you!
Anyway this new rc3 is very reactive, fast! It's always a pleasure to work on Porteus!
Regards,

kalo86

[Blaze]

fanthom, RC3 work fine for me.
I like idea with the tool 'Porteus Packages Manager' from brokenman.
Great job! Thanks.

[Ahau]

Nice work, fanthom!

I've got RC3 up and running and haven't had any issues yet (except an error with the .dat manager, which I couldn't repeat--will test that again today). I haven't had much time to run it through it's paces, and I'm excited especially to use the UUID for my cheatcodes

@ALL:

I have updated the online version of the installation guide, to clean up the bit about MyUAC. I have also added more info regarding bootloader selection in the linux section, and updated all of the NOTES about save.dat and Addendum B, to point folks to use 'Porteus .dat manager' for saving changes on a FAT or NTFS device. Please review the install guide, here: http://porteus.org/info/docs/37-install ... -v-10.html and let me know if there are any errors, inconsistencies, etc., or anything that sounds unclear. Other than our front page, download pages and FAQ, this document is our most viewed article and I want it to be perfect for the new users that will come for V1 FINAL.

I will work on migrating my changes to the .txt file in the ISO, to have the updated doc ready for release in final.

Thanks!

[fanthom]

@kalo86
i can't recreate the error on 2 my PC's. pls provide output of 'psinfo' tool and describe in details how you do reproduce the hang. does it happen straight after boot? did you check if md5sum matches?

@Blaze
good to hear

@Ahau
let's make a deal that all docs which goest with ISO will be ready before Saturday.

############################################
@all
i have made a 'silent' update to rc3 on the server. this is new md5sum:
a647c83cf04fb37f57242f114839fb3c

reason?
1) i have found a bug in 'switch to root' function and plenty of porteus scripts were affected when run from 'guest'.
2) blacklisted rc2800pci kernel module as rt2860sta and rt2870sta still does better job.
3) i have added symlinks from /opt/porteus-scripts/xxx-script to /usr/bin/xxx-script in case one will look for them in /usr/bin
4) good news is that 'Language-Selection-Tool' is finally updated so all non-english speakers can create their own language packs easily

all users affected by rc2800pci bug and all users who use 'guest' by default should download rc3 ISO once again.
sorry for inconvenience.

fanthom

[Ahau]

@fanthom -- Deal!

I have draft updates for all docs except FAQ, will send them over to you tonight for review, subject to changes if anyone finds issues with the current install language on the main site. FAQ will be a little more difficult for me because it's HTML, but I'll get there

EDIT: One more change to install doc (on main site and my local draft) -- added a note that linux users cannot install 64-bit Porteus from a 32-bit linux system, and vice versa.

[82issa]

@Ahau
In the docs under bois config Delete key is a common key for older hardware. Escape key was just for skipping extended memory test on old hardware I don't believe I have never used escape key for entering bios config. Have never used tab before either. I have run into f1,f2,f10 and usually on dells f12 is for changing boot order without messing with bios settings. A onetime boot for the f12.

[Ahau]

Thanks, 82Issa! I remember wondering about that a I was writing it. I'll rework that section.

[Tonio]

I have downloaded this new unplanned release and I see that it works very well.

I have several questions. Why ogg123/oggenc are not on base system? k3b shows the plugins are there, but if I try to convert a ripped cd to ogg vorbis it is not possible without downloading an extra package? Why is this so? I made a module and then I could use oggenc to encode to ogg vorbis and ogg123 to play them/test them. I know mplayer exists, but encoder is not there. I should just add it to this release/final release and be happy

http://www.forum.porteus.org/viewtopic. ... 2f43414507

bash-4.1$ ogg123
bash: ogg123: command not found
bash-4.1$ uname -r
2.6.38.8-porteus
bash-4.1$ uname -a
Linux porteus 2.6.38.8-porteus #1 SMP PREEMPT Fri Jun 10 07:47:57 UTC 2011 x86_64 Intel(R) Core(TM)2 Quad CPU Q6600 @ 2.40GHz GenuineIntel GNU/Linux
bash-4.1$

I did find that the other i486 cds by brokenman & fanthom do have oggenc in the system and one can encode to ogg vorbis there.

I did not use noeject cheatcode:
- 'noeject' cheatcode has one additional feature when used together with copy2ram: prevents unmounting of booting partition

* copy2ram does not eject the cd like it used to , did something change here? had to
$ eject /dev/sr0

[fanthom]

@Ahau
i would like to add docs at the very end in case something changes in next few days.
please send me everything on Friday.

@Tonio
"k3b shows the plugins are there, but if I try to convert a ripped cd to ogg vorbis it is not possible without downloading an extra package?"
it was possible when the last time i was checking it. gonna do it once again to be sure.

"I know mplayer exists, but encoder is not there."
mencoder is there.

"I should just add it to this release/final release and be happy "
ok will add it

"copy2ram does not eject the cd like it used to"
i think it's fixed now. please try this initrd.xz and report back before Sunday:
http://www.4shared.com/file/rBYVswfB/initrd.html

Cheers

[Tonio]

@Tonio
"k3b shows the plugins are there, but if I try to convert a ripped cd to ogg vorbis it is not possible without downloading an extra package?"
it was possible when the last time i was checking it. gonna do it once again to be sure.

"I know mplayer exists, but encoder is not there."
mencoder is there.

"I should just add it to this release/final release and be happy "
ok will add it

"copy2ram does not eject the cd like it used to"
i think it's fixed now. please try this initrd.xz and report back before Sunday:
http://www.4shared.com/file/rBYVswfB/initrd.html

Cheers

not mencoder :>( but oggenc I had to add to download http://pkgs.org/slackware-13.37/slackwa ... 1.txz.html to be able to oggenc some files. I used porteus to bail me out since I had some mp3's and had a screeching sound in Fedora , I could listen to the mp3s correctly with porteus and windows, but they did not sound right in Fedora 15 so I got the idea to convert them to ogg in i486 version it is there

bash-4.1$ ogg
ogg123 oggdec oggenc ogginfo
bash-4.1$ uname -r
2.6.38.8-porteus
bash-4.1$ uname -a
Linux porteus 2.6.38.8-porteus #1 SMP PREEMPT Tue Jun 14 07:33:06 UTC 2011 i686 AMD Phenom(tm) 8250e Triple-Core Processor AuthenticAMD GNU/Linux
bash-4.1$

but in 64 bit, which I will attach later they are not , also in /var/log/packages/libvorbis* package is there in i486
bash-4.1$ ls /var/log/packages/libvorbis-1.3.2-i486-1
/var/log/packages/libvorbis-1.3.2-i486-1


but not in x86_64

I will add an edit here.

Regards,

Tonio

/* Correction, the package appears in /var/log/packages/ but not available for execution

root@porteus:~# ls /var/log/packages/libv* -l
-rw-r--r-- 1 root root 1439 Apr 28 17:15 /var/log/packages/libva-0.32.0_p2-x86_64-1ftm
-rw-r--r-- 1 root root 862 Apr 28 17:15 /var/log/packages/libvdpau-0.4.1-x86_64-1ftm
-rw-r--r-- 1 root root 1297 Apr 28 17:12 /var/log/packages/libvncserver-0.9.7-x86_64-2
-rw-r--r-- 1 root root 9019 Apr 28 17:08 /var/log/packages/libvorbis-1.3.2-x86_64-1
root@porteus:~# which oggenc
which: no oggenc in (/usr/local/sbin:/usr/sbin:/sbin:/usr/local/bin:/usr/bin:/bin:/usr/games:/opt/porteus-scripts:/usr/lib64/qt/bin)
root@porteus:~# which ogg123
which: no ogg123 in (/usr/local/sbin:/usr/sbin:/sbin:/usr/local/bin:/usr/bin:/bin:/usr/games:/opt/porteus-scripts:/usr/lib64/qt/bin)
root@porteus:~# uname -a
Linux porteus 2.6.38.8-porteus #1 SMP PREEMPT Fri Jun 10 07:47:57 UTC 2011 x86_64 Intel(R) Core(TM)2 Quad CPU Q6600 @ 2.40GHz GenuineIntel GNU/Linux
root@porteus:~# cat /etc/porteus-version
Porteus-v1.0

/* new edit

/* edit
@fanthom/brokenman

# tgz2xzm is not available in both i486 & x86_64 bit versions
bash-4.1# tgz2xzm
bash: tgz2xzm: command not found

I think it would be beneficial to have this. I tried to convert wvstreams and wvdial packages created by slackbuilds to xzm and I can't because command is not available

Thanks for your consideration

[jcuk]

Am I in time to report some Problems since RC1? I am entirely "Save to RAM" + VirtualBox module focussed user

First stage testing of RC3, first run with 15+Hours uptime of running Host+Guest VMs.

Having considerable difficulty "passing through" my USB storage devices to any Guest Virtual Machine since RC1
My reference SLAX remix / VB 3.1.6 setup on same hardware is behaving normally.

Had less of a fight this time than with RC2, but its still a huge deterrent to my spending more time running
Porteus x86_64 v1.0 routinely.

Obviously something has changed to affect this since RC1 + VB4.0.x modules released. I had other issues then,
but at least the USB pass-through seemed to work normally. Even in LXDE I could work efficiently, running an
essential main USB drive, and using a variety of USB Flash devices on and off, easily.

With a root KDE desktop: When I plug a storage device in, the "Pop-Up" usually appears in focus over the running
full-screen Guest OS. Usually ( with RC1 ) I just dismissed it, put the Guest Window back in focus, waited a bit,
then went to "Devices" and ticked the new device that appeared. I had a little more of a fight in root LXDE, but did
eventually succeed in umounting the devices from /media (usually from root command line). They seemed to
"automount" and lose information/connection with the GUI Panel widgets for external device state control in RC2

The LXDE / RC2 issue prevented me from doing any significant testing ( Sorry I didnt report it then ) But I got Busy.

I do keep all the logs - from /var/log Host and the VirtualBox Machine. But I cannot figure out what has changed
since RC1 to make RC2 and RC3 so nasty to work with routinely ( and get some useful work done in the VMs with
external storage mounted. The most alarming aspect was a (probably bogus) data/drive corruption warning with
a large xfs archive drive I can ill afford to lose. So I went back to my trusty Slax-remix06 / VB3.1.6 setup until
I saw RC3 on ponce.cc last Sunday.

Uptime on this first RC3 Run now 15Hrs+ My routing survived an overnight turn-off of mobile internet kppp; I
did not need to restart VirtualBox and the Guest OS to regain routing early this morning!

(Root) KDE Desktop quite happily started the two VirtualBox modules; nothing bad happened (yet)
Next: I shall get the appropriate SlackWare packages for aide, gpg, and gkrellm: installing them from a
command line ( separate ) console only login, and do a signature check on this machines Installed Windows
Partitions... If that works out ( and the Host KDE Desktop plus running VMs stay happy ) I'd say I'm closer
to migrating away from last years beautifully stable SLAX-remix + VB 3.1.6

OK! What do I need to do to narrow down this problem? Short of placing my precious xfs main USB drive
integrity at risk! This has to stay mounted for my main dm_crypt User partition for the Ubuntu VM Guest OS.
I can umount it and go entirely native from other User Accounts as needed. But I cant get any work done
in that state - only eating quality time purely testing this Porteus RC3 session.

First - can anyone else reproduce the problems passing-through USB storage devices to Guest VM Appliances?

We are indeed much closer to a system that I can use for routine daily work with some confidence...

Thanks Guys

Posted after 1 hour 28 minutes 26 seconds:
Some important setup detail about how I configured my Save to Ram session ( for those that try to reproduce my USB pass-thru problem ):

Cheatcode adjust (Tab) at "Save to RAM": add ramsize=18% [ 4GiByte Main Memory ]
Set Passwords Both accounts
Logout Guest Desktop immediately
Login KDE root Desktop.
Removed all but US+UK language choices. Set GB as default
Set Monitor Size 1280x1024. Accept configuration ( How do I make this persist across further logouts/logins? )
Setup Dolphin to see Hidden files and detailed list, sort by date.
Go to local disk directory [ /mnt/sda<n>/Porteus64bitRC3 ] where modules are

Raise Command Window. Set a sensible High Contrast default ( not the less visible eyestraining grey fore/black back )
Copy local HDD path in Dolphin URL window
<Paste it into> ln -s <path>/VirtualBox .VirtualBox ( I don't set up Environment Variables, which is probably better for VirtualBox 4.0.x )

OK Now I right click on each of the module names; use the recommended "add module" option from menu.
Back to command window.
cd <paste HDD path again>
command to add extension pack.

Copy sundry .mozila reference profile ans kpprc files to appropriate /root places.

Off we go...

I'm wondering if there is some remaining issue with memory allocation within KDE/LXDE that's causing the problems.
Not tried using a Ctrl-Alt-F1 genuine root console (Only) to add all the extra stuff before running the Desktops. (yet)
( Did this as a successful workaround in RC1/RC2 to add aide + gpg slackware packages then worked from LXDE )

Until weird things happened with USB storage device automounting in the Host Desktop / pass through to the VB module in RC2

[fanthom]

@jcuk
did you install Vbox extension pack as per this thread announcement?
http://porteus.org/forum/viewtopic.php?f=48&t=187&p=841

[jcuk]

@jcuk
did you install Vbox extension pack as per this thread announcement?
http://porteus.org/forum/viewtopic.php?f=48&t=187&p=841

Yup! I omitted to say in my original post that I had, then appended a general statement I'd installed it But not How.
Heres the How
Doing this exactly the same way as for all the previous test 1.0 RC candidates:

From a command window in a root KDE Desktop:

bash-4.1# cd /mnt/sda13/Porteus-v1rc3-x86_64_vb408-T180/

bash-4.1# VBoxManage extpack install `ls -1 *extpack`
0%...10%...20%...30%...40%...50%...60%...70%...80%...90%...100%
Successfully installed "Oracle VM VirtualBox Extension Pack".

As you can see, I'm very careful to keep things in separate directories. So I did use the correct v4.0.8 extension pack.
There are no other extension packs listed and the backtick listing shortcut works nicely in all RC1 runs that I've done.
Actually found for the later versions of v4.0.x (in RC1) that my Guest VMs refuse to start without it loaded.
(4.0.0 did permit it to run without the extension pack, even with the box for extra USB 2.0 support checked.)

Here is the md5sum of the extension pack. Installs OK, so I dont think its a corrupted file. Does this match yours?

bash-4.1# md5sum -b Oracle_VM_VirtualBox_Extension_Pack-4.0.8-71778.vbox-extpack
93882fdb0e808f6ac50ce2b399477b83 *Oracle_VM_VirtualBox_Extension_Pack-4.0.8-71778.vbox-extpack

The Hardware on this machine: 4 * USB rear panel sockets, connected.
1 * USB socket on the panel of a card reader device.
never need to check the "USB Reader" Device to use this USB socket. (as you might expect)

When I tried shifting some data to a USB storage device yesterday, I couldn't get an extra storage device listed until:
I checked that "USB Reader" device box
Presumably that forced a rescan, and only then did it find the device. Even though it was plugged in to a rear USB socket
all along.

There seems to be an issue with device scanning and communicating those changes when working with VirtualBox.

But I dont know if its the fault of the VirtualBox interprocess communication with the Host Desktop or something
fundamentally wrong with the host Desktop(s). I've had similar troubles with LXDE in RC2. Something changed after RC1.

System is still useful, and running healthy. Uptime 30+Hrs, and another long absence of kppp running overnight didnt
lose routing when I restarted it this morning. After the Guest VM morning cron job.

Happy to try some more experiments, and copy logfile information / describe what happens. After a fight to get my
main USB external HDD recognized and mounted in the Guest VM I left it mounted. Need it to access my main user account.

What testing/inspection do you recommend that would help to isolate the cause?

[fanthom]

hello jcuk,

I was able to start Virtualbox in Porteus (KDE desktop) and assign USB stick to Porteus virtual session. All data were transferred successfully from virtual session to the stick.

I wasn't able to do it in LXDE. Seems that LXDE/udisk automount support is broken in 32bit Porteus. At the very start i see 'lxpolkit segfault' in /var/log/dmesg.

Code: Select all

lxpolkit[3903]: segfault at 8 ip 08049ba4 sp bfda7cc0 error 4 in lxpolkit[8048000+3000]

I need to find if this is caused by Porteus itself (config? overriding files?) or LXDE (lxpolkit in particular) must be recompiled on Porteus. Also need to check if 64bit version is affected by this bug.

Stay tuned for news.