You are not logged in.
Hello:
I think the clue is in the error printout:
timeshift: Timeshift cant restore RSYNC snapshots if your system installed on BTRFS with wrong subvolume
See here: https://forum.manjaro.org/t/possible-so … rfs/149901
Best,
A.
Hello:
... will be utilizing microcode updates to protect myself ...
As far as I know, it is enough to keep your system up to date.
In the last nine years, have seen intel-microcode packages updated once every blue moon, probably because my box runs on a legacy (EOL 03/2013) Intel Yorkfield (Core™2 Quad Q9550) processor.
The intel-microcode package is/has been there as part of the installation/upgrades from the start:
~$ apt list | grep installed | grep -i microcode
--- snip ---
intel-microcode/oldoldstable-security,now 3.20231114.1~deb10u1 amd64 [installed]
~$
And then there is what you can see here.
Microcode for Intel and amd64 CPUs all the way from up Jesse down to Ceres.
So any time a microcode package gets upgraded, it is made available for you in the Devuan repositories.
You would have to take intentional steps to keep your system apt from actually downloading and installing it.
I may be missing something in your particular case, but in my opinion, keeping your system up-to-date is not an option to consider.
It is something you do.
Best,
A.
senHello:
... set sudo to execute certain commands ...
Indeed.
But I was wanting to avoid adding to my overpopulated sudoers.d directory and thought that maybe there was a way to get that done within conky for cases like this.
Guess that will have to do. 8^)
I'll test it and see how it goes.
Edit: that did it - works as advertised.
Thank you very much for your input.
Best,
A.
Hello:
After much procrastinating, I finally set about cleaning up my conky.conf file.
This year's has been a very hot summer and the fans in the box many times ran at high speeds.
I need to know what is going on inside temperature wise.
The on-board w83627ehf chip is not activated and loading the module gives bogus/unreliable readings.
The only way I can get a reasonably approximate in-box temperature reading are from the pair of SCSI cards I have inside, their readings should be good enough to get an idea of internal temperatures.
So, to that effect, I need to run these lines in conky:
Case 0: +${exec sensors drivetemp-scsi-5-0 | grep 'temp1' | cut -c 30-39} C
Case 1: +${exec sensors drivetemp-scsi-3-0 | grep 'temp1' | cut -c 30-39} C}
The problem I have is that sensors can only be run as root so it is not working from conky.
~$ sensors drivetemp-scsi-5-0 | grep 'temp1' | cut -c 30-39
Error: File /etc/sensors.d/sun_u24.conf: Permission denied
sensors_init: General parse error
~$
~$ sudo sensors drivetemp-scsi-5-0 | grep 'temp1' | cut -c 30-39
[sudo] password for groucho:
+34.0 C
~$
Everything else works properly.
Any pointers will be appreciated.
Best,
A.
Hello:
@czeekaj
For wireless I'd use this.
Had a look, will put it into the list of possible replacements.
@pcalvert
Did you try connman ...
No, not yet.
The thing is that I am still on the fence as to whether to go through with the Beowulf -> Chimaera, among other things because I am not sure I want to suffer the new version of xfce4, more so with the upgrade itself being rather problematic.
I think I'll wait this transition out for a while longer.
Thank you both for your input, much appreciated. 8^)
Best,
A.
Hello:
Disable os-prober before you do any upgrading.
It's been default disabled.
I think grub-customizer probably uses it.
But I have not used that application in more a good long while, when I had another HDD with another version to test.
That said, when/if upgrading, the idea is to clone the system drive, set the box to boot from the clone and only then do any upgrade.
Then, if there are no issues, make an image of that drive to restore back to the system drive.
As this is a BIOS boot (no UEFI stuff) box, any other OS boot issues generated by the OS-prober can be easily fixed there.
... update critical packages first ...
If I decide to continue with attempting this upgrade, the system drive will be fully updated and cleaned of left over cruft/orphans before cloning it.
... make sure you have an up to date keyring ...
Thanks for reminding me.
Seems it was up to date because during the experiment (see part II, link below) there were no warnings in the terminal printout.
And I accepted all the default options when given the choice.
Not sure that was a good thing as I once had a severe problem with the default Devuan MTA, Exim4 by doing just that.
ie: accept default suggestions when updating/upgrading.
TL;DR: I removed/purged it and in its place installed a lightweight and much better maintained one called procmail.
After all, I was only needing a local MTA, nothing too elaborate.
Just works and in three years' use it has never given me any grief, to the extent that I had to actually look up the package name. 8^°
Thanks for your input.
Best,
A.
Note: I will mark this as SOLVED to this thread will continue in part II here.
Hello:
... already demonstrated to you what wicd works in Dedaulus.
Yes, I remember.
Thanks for that. 8^)
But now the problem at hand is first being able to make a painless move from Beowulf to Chimaera.
Incredibly enough, there were no issues with wine, my MS based Pegasus Mail email client or the unsupported legacy Nvidia 340.108 drivers.
Upgrading to Daedalus is not in my sights at this moment.
Whatever for? <- which remains an unanswered question somewhere in my OP.
My WiCD setup/configuration did not survive the upgrade, probably due to (surprise, surprise!) xfce4.
I use WiCD to switch on/off my eth0 connection with just a click as well as being able to get up and running (via my friendly neighbour's WiFi) in under 5' should something happen with my ADSL line, the router or the braindead AHs who run the local telco improvement and modernising departments.
So, at least for me, it is not just for WiFi.
I may have to switch to MATE or directly to Openbox but the lack of an easy to use configuration tool is a bit of an issue for me.
I seem to recall (?) that there was a script (somewhere) which would generate a #! type desktop without much ado.
Can't remember now, maybe it was related to something Hoas wrote up.
Thanks for your input.
Best,
A.
Hello:
... Chimaera has wicd.
Hmm ...
Not really.
Have a look here.
Origin: Devuan
Description: wired and wireless network manager - transitional package
That is a WiCD package only by name.
It is a transitional package that works as a front end to install any one of its listed dependencies:
ie: network-manager-gnome | network-manager | connman-gtk | cmst | connman-ui
From the README.Devuan file:
Wicd was removed from Debian Bullseye as it requires python 2.
This transitional package ensures that users upgrading from Devuan Beowulf get an alternative* network manager installed. It can safely be removed.
* underline is mine.
Thanks for your input.
Best,
A.
Hello:
This is a follow up on this post.
More than anything to avoid excessive length and muddling up.
To make a first, quick test run, I booted up my box via a Daedalus 5.0 desktop-live *.iso and and generated an *.img file of my system drive.
That done, I restored that *.img file to an empty, larger drive sitting in one of the cage sleds.
That worked prefectly well, the restored image taking up its space leaving unallocated space untouched.
Being a test, I chanced using gnome disks which surprised me, but then this is a vintage/non-UEFI rig.
I then set up the BIOS to boot from the drive where the newly restored image was residing so that the system would not touch it.
Once I rebooted the box, I checked and found that everything was working as expected and nothing seemed to be amiss.
A bit slower as my system drive is a Samsung 120Gb SSD and this is just an older 300Gb SAS drive.
I then followed the instructions on this page at Dev1.
Clear and straightforward enough, as always/usual.
Being on a WiFi connection, the whole thing took around 45/60 minutes or so, didn't bother to time it.
But after rebooting I found that my xfce desktop settings had all gone south, the panels were empty and I no longer had WiFi access.
ie: I had no way (interface) of making the bloated POS network-manager work.
It seems that the xfce upgrade signifficanly screws up most if not all previous settings.
Why am I not at all surpridsed?
Eventually and after much trial and error, I managed to bring up a WiFi connection via command line to complete something that had not been updated/upgraded correctly.
But after rebooting even nmtui refused to connect to the very same WiFi I could instantly connect to with WiCD/Devuan Beowulf in my Asus 1000HE netbook. Yes, all credentials were correct and I have always been able to connected both machines to the router at the same time.
That said, I think that if I want to ge this right, the first thing I need to do is find a proper replacement for WiCD.
ie: one that works like WiCD, that I can bring up and use to connect to whatever I need/want to use (wired or wlan) without any hassle.
Just like I can do today with WiCD in Devuan Beowulf.
On the bright side of things, SLiM worked as expected, my screen setup worked as usual (nvidia drivers) and even wine brought up my email client (Pegasus Mail) with no issues.
But xfce and network-manager absolutely screwed up, ruining the show.
I am now back to my original system drive thinking about what to do.
Any comments will be welcome.
Thanks in advance,
Best,
A.
BTW: writing up this post took me a while, as a result, the timeout no one else suffers from kicked in. Still don't know why or how this happens.
Hello:
... have posted information about some efficient methods here ...
Thank you very much.
I'll have a look at all that.
Best,
A.
Hello:
... look at my 2 dev1galaxy cookies, no luck.
... tool that allows to read the contents ...
Same here.
... is there a tool within FF that allows to read the contents?
I don't think so, at least I have not found it.
If my memory serves me right (?) a long time ago, opening the settings page in a browser (cannot recall which one) you could see some cookie properties, one of them being their expiration date. I recall having seen cookies with a years' long expiration date. (!)
Have not tried it, but see here.
Bear in mind that it is from 5 years ago so I don't know if it still applies.
Best,
A.
Hello:
... not aware of any timeout related to the login time ...
... stay logged in as long as I like.
I see.
... FF ESR latest version on Daedalus.
FF 115.8.0esr (64-bit) on Devuan Beowulf here.
... related to a forced IP change of your internet provider?
Don't think so.
My friendly local telco providing me with a landline since 1986 along with a pricey albeit low quality ADSL for the past five years decided manu militari to make good on their advertised improvements and deactivated it. With no previous notice.
I am now a 2.0 USB WiFi dongle, legitimately loaned by a neighbour two doors down the hallway so it is not that, the IP is always 192.168.1.29 and an old tin can antenna makes for a very decent 85/90% signal. A good temporary arrangement till the telco AHS deign themselves to come and see about a installing fibre connection.
But this also happened when my ADSL was working through a Pi-hole/Unbound rig running on a Chimaera VM inside my box, the IP also always 192.168.1.2 and the DNS 192.168.1.5. Worked quite well.
I think that the login credential is on a Dev1 cookie and probably has an expiration date.
eg: at this moment I have two Dev1 cookies in my browser but they show no expiration time/date which I recall could be seen in older browsers (long time ago).
Maybe if I change the setting from Allow for Session to Allow this will stop happening?
I always set any and all browsers I use to clear all history on shutdown.
Thanks for your input.
Best,
A.
Hello:
When answering a post, whether it is long or needs a longish reply (+taking the time to edit/crop/spell appropiately or maybe look up something like the proper spelling of a difficult word) more than once I find that when I hit 'preview' I have been logged out.
ie: some system set timeout has lapsed, no idea how long it is.
In all these years at Dev1 I have 'sort of' become used to it and just log into Dev1 in another tab and go back to the original one.
Hitting 'preview' again in the original tab will make the browser find the Dev1 cookie and I can keep go ahead and edit or post.
But it does not always work as intended and ended up losing finished posts only to have to write them up all over again.
eg: again this morning ... 8^°
Fortunately, my short term memory still functions properly, but still, not all is always remembered. 8^°
I don't know how long that timeout is but for how I post ie: trying to be concise, clear, tidy and on an easy to look at page, it ends up being short.
Others may opine that it is a workflow problem that could be solved using an external editor to write a draft.
But no, it is not a workflow problem: I care about how and what I write and go over my text a few times before I post it.
I consider not posting in haste or 'off the cuff' a matter of respect to my fellow Dev1 members
I also consider that using an external editor would totally defeat the whole idea of using the browser as an editor to post on a site like Dev1.
Some sites automatically save a draft for a few hours/days/permanently when something like this happens.
For whatever motive or if the the poster's system/connection goes down.
So ...
I was wondering if it would it be possible for the Dev1 admins to consider extending that timeout to maybe 50% longer?
Thanks in advance.
Best,
A.
Hello:
... only way to be sure is to send it.
Always done that.
I started with Devuan Jesse in this box and from then on up to Beowulf (now with a backported kernel) was like that.
I opted out of Chimaera because of WiCD and SLiM, this last one having been revived.
But the 'Debian Way' ® is not what it was, between the bloat, the infamous usr/merge and who knows what else is lurking around there, I am rather wary.
Who knows what sending it may bring about.
I will probably clone my system 120Gb SSD drive to an empty 300Gb SAS I have in the slots.
Boot from there and see what happens when I send it and reboot.
Cannot wait to see my dmesg printout. 8^D
That said, I have noticed that what I see as the most important part of my screed has not gathered the much attention.
ie:
Seeing that my hardware has not changed significantly since I first installed Devuan Jesse ca. 2017 and will most probably not change in the coming years, is a dist-upgrade necessary? This is not a file-server exposed to the web.
Thanks a lot for your input.
Best,
A.
Hello:
... prefer to use Clonezilla.
Refracta is a good choice for this.
... also make a second backup using FSArchiver.
Right.
... separate additional backups of /etc and /home ...
Yes, BackInTime takes care of my /home directory and Timeshift takes care of /etc so that is covered.
re: gnome-disk-utility
I had my doubts which is why I asked.
Seeing that no one among 400+ visits answered with a clearly pronounced yes I won't be using it.
Thanks for your input.
Best,
A.
Hello:
I assume you are using Nvidia drivers.
If so, RandR may/will not work along with Xinerama.
I use a pair of FX580 cards for three 19" monitors with the (now unsupported) Nvidia 340.108 drivers and an xorg.conf file I managed to put together after a very long trial and error process, obviously with Xinerama enabled.
At one point, on every boot my /var/log/Xorg.0.log file reads:
~$ cat /var/log/Xorg.0.log
--- snip ---
[ 91.365] (WW) NVIDIA: The Composite and Xinerama extensions are both enabled, which
[ 91.365] (WW) NVIDIA: is an unsupported configuration. The driver will continue
[ 91.365] (WW) NVIDIA: to load, but may behave strangely.
[ 91.365] (WW) NVIDIA: Xinerama is enabled, so RandR has likely been disabled by the # <---- !!!
[ 91.365] (WW) NVIDIA: X server.
--- snip ---
I believe this is applicable to all Nvidia drivers (unsupported or not) but I am absolutely not sure about that.
That said, I have recently seen that the nouveau driver works quite well (daedalus-live desktop *.iso).
You may want to consider trying out the live daedalus *.iso and see for yourself.
I was quite surprised to see how well it worked and how easily I could make the monitors behave as I wanted.
Best,
A.
Hello:
... if the Ethernet cable going bad ...
... only way to test it is to change the cable.
In my experience, a good quality ethernet cable (of the right spec, obviously) is essential to achieve and maintain advertised speeds.
It is also good practise to keep a couple of unused ones handy just in case.
Those tiny/crappy plastic tabs thingies to keep them in place are a PITA.
Best,
A.
Hello:
... don't take my words as a reproach ...
Of course, you have not given me any reason to do so.
... “sat through” on opensuse 12.xx almost two years ...
I can relate to that.
The coffee roasting software I use runs in my 32bit Asus 1000HE under Devuan Chimaera.
It is the last 32bit version (1.1 / ca. 2017) made available by the author works perfectly well with all the hardware involved.
All versions after that (v2.10) slowly acquired a huge amount of professional roasting machine bloat.
That same machine has an XPSP3 partiiton I use once in a blue moon.
It came in very handy when I needed to flash the last available OS on a Samsung G530M now unsupported by the local telco.
For whatever reason my local XP VM would not run the MS based software I needed to get the job done without screwing up half way through the process.
... replacing it turned out to be practically a new installation.
Yes, I can relate to that too.
... at some point we will have to change the OS.
Indeed ...
I foresee having bad dreams about that.
There's not really much to choose from if Devuan should fail.
A Devuan based Linux from Scratch may very well be waiting for us in our collective future.
Thanks a lot for your input.
Best,
A.
Hello:
re OP and re HARDWARE... nice!
... might enjoy this ...
Thank you.
Will have a look.
... under the consideration that at any moment ...
I have no problems with the SSD.
fstrim, BackInTime and Timeshift are always there to watch my back.
Thanks for your input.
Best,
A.
Hello:
... play around and then you will understand ...
I am afraid that I may not have explained myself properly, again.
The only thing bothering me is the idea that, given the reasons I tried to explain in my OP, an upgrade from Beowulf (with a backported kernel) to Chimaera and eventually Daedalus may not be worth the trouble and may even be counterproductive.
No one will decide this for you.
Goes without saying. 8^°
... rest is empty talk.
I beg to differ, strongly.
As always, YMMV.
I am not looking for a decision to be made for me.
What I am looking for are opinions/insight to be able to make a better and more informed decision.
I learned a (very) long time ago that new, latest, shiny and full of bling do not necessarily make for a better product.
The IT world (OSs, software and hardware) is up to the brim with examples.
Thanks for your input.
Best,
A.
Hello:
You determine ...
... happy with the OS and the software ...
... keep Beowulf.
That is more or less what I think.
... no more upgrades/safety fixes ...
Can you live with that?
Really cannot say.
... seen your posts about xorg security issues.
I posted them as informative for the forum.
And my Beowulf system received them within 24 hours.
That said, I never noticed any issues with xorg.
... removal of support for older hardware.
Quite so.
A few years ago I eventually had to upgrade all my hardware because there were no Linux drivers for my discontinued but prefectly working (and very expensive) Matrox G450 cards.
... installation with --no-install-recommends ...
Yes, I have apt set up to do just that.
Thanks for your input.
Best,
A.
Hello:
... assume its a legacy installation ...
Aye, it is.
None of that UEFI stuff in my box or in my netbook.
... switching drives should be painless.
I am afraid that I may not have explained myself properly.
The system SDD is just fine, no problems with that.
My issues are with the move from Beowulf to Chimaera and then (eventually) Daedalus.
@aluma
@Andre4freedom
@rolfie
Independent of the method used to make a back-up image of my system drive and for the reasons stated in my OP, the question remains:
Seeing that my hardware has not changed significantly since I first installed Devuan Jesse ca. 2017 and will most probably not change in the coming years, is a dist-upgrade necessary?
All I see around me these days is bloat, bloat and more bloat.
To what benefit?
Thanks in advance.
Best,
A.
Hello:
... post a file with information ...
Check with golinux.
A sticky would be a good idea provided it passes review. 8^D
Gnome-dosk-utility might work ...
In this case, might is a scary term. (even more so than dosk) 8^°
I need to go through the dist-upgrade routine and be certain I can go back without any hassle if something gets screwed up in the process.
Thanks for your input.
Best,
A.
Hello:
... no need to break a working system.
Indeed ...
That is exactly what I am trying to avoid, as well as a clean install.
... one partition "/" on which to install Daedalus 5.0.
As I noted above, I run Beowulf with a backported kernel.
~$ uname -a
Linux devuan 5.10.0-0.deb10.16-amd64 #1 SMP Debian 5.10.127-2~bpo10+1 (2022-07-28) x86_64 GNU/Linux
~$
I understand that I should not skip Chimaera before upgrading to Daedalus.
... a live gparted image is better.
Right.
Note taken.
Thanks for your input.
Best,
A.
Hello:
My Devuan Beowulf (backported kernel) is now long in the tooth and I am rather undecided with respect to moving on to Chimaera.
The main reasons being that everything is working quite well and that the hardware in my box has not changed since I first installed Devuan ca. 2017.
It is a ca. 2007 Sun Ultra 24 workstation, nicely loaded with everything I need, including my trusty Umax S-6E SCSI scanner running on an Adaptec 2940UW PCI controller card which at one time also served a pair of 4.0Gb DAT drives.
Save issues with one of my monitors or HDDs, I do not expect any hardware changes in the next four or five years.
Making sense?
Or is my atavistic aversion to change getting the best of me?
There is also the issue of all the work I have put into configuring everything, the sole prospect of having to do it again trumps any idea of a clean install so, if I were to decide to go ahead with the move to Chimaera, it would have to be a dist-upgrade.
If I were to go ahead, I will need to make a foolproof image of my system drive, a Kingston 120Gb SSD.
In another life, Norton Ghost saved me a lot of work while doing MS/PC maintenance at a ministry full of users who knew a lot about computing.
Is disks ie: gnome-disk-utility 3.30.2 up to the task?
Thanks in advance.
Best,
A.