- main - earth.li

Re: Server email with minimal overhead
by Phil Ashby 20 Sep '22

20 Sep '22

On 15/09/2022 12:00, Mark Rogers <mark(a)more-solutions.co.uk> wrote: > > I always seem to struggle with this most simple of configuration tasks! > > Starting from a clean install of (say) Ubuntu/Debian, and with a set > of SMTP server credentials in your hand, what's the best/simplest way > to get email set up so that emails from cron jobs get sent, and so > that there's a minimal "mail" client at the commandline? > > I don't want the overheads of postfix + MySQL that things like > mailutils bring in. (Or maybe I should accept those overheads?) > > The broader objective is that I can set the recipient address for > email to root or other users, and the sender address from which emails > get sent by default. I would opt for msmtp-mta (standard package) which fakes sendmail: https://packages.debian.org/bullseye/msmtp-mta No daemons, queues, special accounts or other malarky, just an API that other programs know how to operate and a protocol engine that can talk SMTP to your real mail system. P.

2 3

FLAC CD Ripper?
by Steve Mynott 17 Sep '22

17 Sep '22

Can anyone recommend a native Linux FLAC ripper which auto tags music with basic info like band and song name? Maybe something which resembles EAC since I've used EAC under Wine in the past which used a Free CD metadata library, but I'm not sure if the library is available any more? -- Steve Mynott <steve.mynott(a)gmail.com> rsa3072/629FBB91565E591955B5876A79CEFAA4450EBD50

1 0

cmos jumper?
by Phil Thane 30 Aug '22

30 Aug '22

Hi, After years of upgrading and tinkering with my old desktop, I bought a 'new' one. 'New' as in assembled cheaply by my local computer shop from less than state of the art bits for £300. The machine came with a smallish SSD so I've fitted the 1TB HDD from my old PC alongside. Installed Mint on the SSD and linked my Nextcloud directory on the HDD onto my desktop. On the whole it works fine except the boot routine. On boot I get something like 'Unable to find operating system'. If I tap f9 during boot I get a table of boot options and choosing SATA 0 (ie the SSD) makes it boot as expected. But obviously it's annoying. f10 (I think) takes me into BIOS settings and it appears I can set boot order something I've done on various machines for donkey's years but on this one it doesn't stay in memory, on the next boot it once again cannot find the OS. Also, I noticed the date/time settings keep going wrong. Usual cause of this is either corrupted CMOS settings or more likely a faulty mainboard battery. Either could also be preventing BIOS storing settings. So off with the cover and look for the CMOS jumper. Can't find it. Can't find a decent hi res image of the board online and the instructions on the board manual tell you to move the jumper from pins 1 + 2 to 2+3 (which I knew) but not where it is. Near the battery is a 2 pin header with no label (could be a fan?) and a bit further away a multipin header but I can't make out what that's for, possibly USB? neither has any connectors or jumpers on. So, next step, I removed the battery and tested it. 3.12V which is fine. Left out for a few minutes which should reset BIOS but it didn't cure the problem. I've spent ages searching for info on this specific board with no luck. Can anyone help? Details: Mainboard MS-7778 VER 1.0 On boot I get an HP logo so I suspect the shop recycled it from an HP machine. -- Phil

2 1

Re: How to Change Group
by Phil Ashby 14 Jul '22

14 Jul '22

On 13/07/2022 12:00, steve-ALUG(a)hst.me.uk wrote: > > Hi All > > Due to some historical ineptitude, I have ended up with this situation. > > On my Server, I have UserA, with User ID 1001, and Group ID 1001. > > On my main laptop, I have > > User A UID 1001 GID 1002 > User B UID 1002 GID 1001 > > I've just discovered this. Now, when I open files on a NFS share, > they're shown on my laptop as owned by User A, Group User B, despite the > server thinking they're both User A. > > On my laptop, I need to change the group IDs round in the /etc/group > file so they're consistent with the user IDs. I also need to change any > files legitimately owned by User A or User B to have the "correct" group ID. > > User A is my main laptop user. I have got another - User C I can login as. > > I suspect, what I'll have to do is > > Login as User C > > #something like... > > #change files gid 1001 to gid 4444 > sudo find {magic to find Gid 1001} {exec's strange syntax to} chmod > :4444 "the file" > > #change files gid 1002 to gid 1001 > sudo find {magic to find Gid 1002} {exec's strange syntax to} chmod > :1001 "the file" > > #change files gid 4444 to gid 1002 > sudo find {magic to find Gid 4444} {exec's strange syntax to} chmod > :1002 "the file" > > Then I'll need to change the GID numbers somehow in the Group file. The > simplest way I can think of is just edit the file, save it, then > immediately reboot. > > Before I do this, can anyone see any GOTCHA!s? > > Any suggestions or advice please? Hi Steve, I've had to do similar a few times (also ineptitude in the past, and adding different people to different machines, while having an NFS store in the middle). Gotchas: - everything will be weird until the users log off/on again. - you should re-mount the NFS to drop caches (or restart machine). - files on the server need processing too if there are any from userB Process: - as suggested, go via a non-affected user C, do the sudo/su dance as appropriate for your setup. - use 'vigr' (and 'vigr -s') to update the group files so that you have the IDs allocated as desired (using vigr tell the kernel to re-cache) - unless you know there are files outside user home folders (unlikely), do not attempt a find/chgrp, simply apply a recursive change to the home folder # cd ~userA # chgrp -R userA . # cd ~userB # chgrp -R userB . - safest to restart the machine, and check all is as you expect :) - if you want to try a find/chgrp.. here's the magic I know: # find / -mount -group <gname> -print0 |xargs -0 chgrp <newgname> > from the root, without crossing mount points (you might omit this if you're sure the NFS mount isn't there!), find files with the specified group name or number, spit 'em out with \0 separators to avoid pain with white space in file names. Pipe to xargs which expects \0 separators, then runs chgrp to the required group name in large (efficient) blocks. Cheers, P.

4 3

How to Change Group
by steve-ALUG＠hst.me.uk 13 Jul '22

13 Jul '22

Hi All Due to some historical ineptitude, I have ended up with this situation. On my Server, I have UserA, with User ID 1001, and Group ID 1001. On my main laptop, I have User A UID 1001 GID 1002 User B UID 1002 GID 1001 I've just discovered this. Now, when I open files on a NFS share, they're shown on my laptop as owned by User A, Group User B, despite the server thinking they're both User A. On my laptop, I need to change the group IDs round in the /etc/group file so they're consistent with the user IDs. I also need to change any files legitimately owned by User A or User B to have the "correct" group ID. User A is my main laptop user. I have got another - User C I can login as. I suspect, what I'll have to do is Login as User C #something like... #change files gid 1001 to gid 4444 sudo find {magic to find Gid 1001} {exec's strange syntax to} chmod :4444 "the file" #change files gid 1002 to gid 1001 sudo find {magic to find Gid 1002} {exec's strange syntax to} chmod :1001 "the file" #change files gid 4444 to gid 1002 sudo find {magic to find Gid 4444} {exec's strange syntax to} chmod :1002 "the file" Then I'll need to change the GID numbers somehow in the Group file. The simplest way I can think of is just edit the file, save it, then immediately reboot. Before I do this, can anyone see any GOTCHA!s? Any suggestions or advice please? Thanks in advance. Steve

1 0

[OT] Firefox loses cookies periodically
by Bev Nicolson 11 Jul '22

11 Jul '22

I'm on vs 101 I think now, and periodically Firefox will dump the cookies that I have given an exemption to. (I ask it to delete all the others on closing the browser and have a limit on how much disk space the rest can take up.) What can I do about it? Bev. (Mint 19.1)

1 0

Sharing disk space between servers
by Mark Rogers 27 Jun '22

27 Jun '22

I have several virtual servers with the same provider. Typically I have around 50GB disk space. On some I have largish databases, with consequently large database backups. The backups are for recovery from cock-ups rather than disaster recovery so retaining them on the same server is convenient, but they do tie up disk space, and the ones with the biggest databases lose the most to database backups. So ideally I want to do something to share the spare capacity on other servers, but safely. for example in the event that one server got hacked I don't want that to provide a backdoor to another server, and nor should any data that's been shared onto it be easily accessible. In an ideal world, it would be nice to ring fence (say) 15GB of capacity on server A and make it available over the network to server B such that A can't access B's files even though they're stored there (I accept that I can't stop someone with root on A deleting the files, but they should at least be encrypted so that they can't be read). And whilst B can use that allocated chunk of A's disk, it must not be able to access anything else on A. Ideally the capacity that's shared should be pretty easy to change so I can easily adapt if A needs more space. (I wondered about trying to remotely mount an encrypted file on A as a filesystem on B, but adapting the size later would be difficult.) As anything I do will be making changes to already functioning servers A & B the changes need to be relatively narrow to avoid the risk of breaking something. Thoughts / suggestions / experience of doing something like this? I can think of various paths to take, from simply pushing files over scp and working out how best to secure it (and manage the quota), to something far more complex like GlusterFS, with lots of options in between. So it's not really a lack of ideas as any idea which ones would be best! -- Mark Rogers // More Solutions Ltd (Peterborough Office) // 0344 251 1450 Registered in England (0456 0902) 21 Drakes Mews, Milton Keynes, MK8 0ER

1 0

Editing files on remote servers over SSH
by Mark Rogers 26 Jun '22

26 Jun '22

What's the best way to edit files on remote servers over SSH? Using a text editor directly over SSH is fine, but a GUI editor like VSCode is easier to work with. But I never have much luck getting SSH/SFTP plugins to work reliably, so I wondered what other people do. Preferably cross-platform as I have to switch between platforms a lot. -- Mark Rogers // More Solutions Ltd (Peterborough Office) // 0344 251 1450 Registered in England (0456 0902) 21 Drakes Mews, Milton Keynes, MK8 0ER

3 3

SSD HDD and simlinks?
by Phil Thane 01 Jun '22

01 Jun '22

After plaguing this list for months about my ailing desktop PC I finally skipped it and bought a (basic) new one. It comes with 128GB SSD. It's fast of course but not very big. So here's the plan but any advice would be good... I took the (almost new) 1TB drive out of the old PC and there's a lot of stuff on it. I've fitted it to the new machine but not installed anything yet as my ancient keyboard with a PS2 plug won't fit! Once that's sorted I'll install Mint. I *think* I'll do a standard install to the SSD and ignore the HDD. My user data on the HDD is mostly in /home/phil/Nextcloud and it was synced to my Nextcloud server. I assume, hope, I can keep that synced data on the HDD and just link the directory to my new /home/phil/... Not sure if that means using the link as the Nextcloud directory when setting up the client or maybe Nextcloud will allow me to connect direct to the HDD. Either way though I'm hoping to avoid taking all day to sync my data across the network. I remember long ago having a PC with 2 small HDD and setting it up with / on one but /home on the other. Is that still a thing? Is it worth doing in this case? Though bearing in mind it means formatting the HDD and re-syncing all the data. -- Phil

2 1

Re: Comparing large folders of mostly identical file contents with different filenames (Mark Rogers)
by Peter 31 May '22

31 May '22

A query, not a suggestion, because you are all far, far better informed than me about this stuff! I was curious and looked up what tools are available for this, and came on: rdfind Fdupes dupGuru which all seem to find dupes by content. I am sure Mark is well aware of these and more. So why can't one of them be used to find the dupes? Also, is any of them significantly better than the others? Different but related issue, the app I have used to sync folders in the past is Unison, which I found very simple and effective. Also Grsync. Do people use these? Or is something else preferable? Peter On Sun, 22 May 2022 12:00:03 +0100 main-request(a)lists.alug.org.uk wrote: > Send main mailing list submissions to > main(a)lists.alug.org.uk > > To subscribe or unsubscribe via email, send a message with subject or > body 'help' to > main-request(a)lists.alug.org.uk > > You can reach the person managing the list at > main-owner(a)lists.alug.org.uk > > When replying, please edit your Subject line so it is more specific > than "Re: Contents of main digest..."

8 14