Re: [gentoo-user] Strange output when restarting network. Long term issue.
Hello,
On Thu, 23 Jul 2020, Ashley Dixon wrote:
[..]
>[super] root@ad-gentoo-main / # etc/init.d/net.eno1 restart
[..]
>forked to background, child pid 14536
[..]
It's just output from backgrounded processes. Just ignore it or enter
Ctrl-C to get a fresh prompt. Using enter is not recommended.
Compare running:
{ echo "start"; sleep 2; echo "done" ; } &
HTH,
-dnh
--
My word processor was written by Stanford Professor Donald Knuth. Who
wrote yours?
Re: [gentoo-user] Strange output when restarting network. Long term issue.
On Fri, Jul 24, 2020 at 02:59:10AM +0100, Ashley Dixon wrote: > I've just run it again, and it seems like the culprit is in the _show_address > function, which calls einfo [1], passing the result of _get_inet_address. > I'll > have a look at this more now, as it's something I also would like to be > resolved. It seems like this is not the case. This might be a bug in netifrc, or the OpenRC init system helper functions in general. I can't find anything on-line regarding this issue, although I'm sure it's reasonably common. -- Ashley Dixon suugaku.co.uk 2A9A 4117 DA96 D18A 8A7B B0D2 A30E BF25 F290 A8AA signature.asc Description: PGP signature
Re: [gentoo-user] Strange output when restarting network. Long term issue.
Ashley Dixon wrote: > On Thu, Jul 23, 2020 at 08:46:52PM -0500, Dale wrote: >> At least it isn't just me. Sometimes I think bugs/bad features pick on >> me. While annoying, I guess it is harmless. It just seems that as long >> as it has been doing this, someone would have raised the hood and looked >> to see what is causing it and how to fix it. Strange that something >> weird like this has been around so long. > I've just run it again, and it seems like the culprit is in the _show_address > function, which calls einfo [1], passing the result of _get_inet_address. > I'll > have a look at this more now, as it's something I also would like to be > resolved. > Thanks much. Hopefully it is a easy fix. Dale :-) :-) Also tracking your euses package. Got it installed but the patches are giving me grief when I try to update. Hopefully it will hit the tree soon and be updated normally.
Re: [gentoo-user] Strange output when restarting network. Long term issue.
On Thu, Jul 23, 2020 at 08:46:52PM -0500, Dale wrote: > At least it isn't just me. Sometimes I think bugs/bad features pick on > me. While annoying, I guess it is harmless. It just seems that as long > as it has been doing this, someone would have raised the hood and looked > to see what is causing it and how to fix it. Strange that something > weird like this has been around so long. I've just run it again, and it seems like the culprit is in the _show_address function, which calls einfo [1], passing the result of _get_inet_address. I'll have a look at this more now, as it's something I also would like to be resolved. -- Ashley Dixon suugaku.co.uk 2A9A 4117 DA96 D18A 8A7B B0D2 A30E BF25 F290 A8AA signature.asc Description: PGP signature
Re: [gentoo-user] Youtube-dl and file time stamps.
Simon Thelen wrote: > [2020-07-15 17:30] Dale >> Howdy, > Hi, > >> I'm not sure what causes this because it doesn't always do this. When I >> use youtube-dl to download videos, it sometimes uses the current date and >> time for the time stamp. I like that because I can sort by date and see >> new videos. On some sites tho it seems to use the time stamp of the file >> on the server I am downloading from not when it was put on my system. >> Sometimes I download a video and it may have a time stamp of years ago, >> decades sometimes. I looked through the help page but can't find a option >> to tell it to use local time instead of the time from the remote server >> file. Needless to say, when it does this, I can't tell which videos I >> recently downloaded since sorting by time stamps is no longer accurate. >> It's annoying. >> >> Has anyone else noticed this behavior? Is there a way to tell it to stop >> setting it to really old time stamps? Some option that isn't documented >> maybe. > You're probably looking for the --no-mtime option. Depending on what > you're using to sort your local videos you can always just tell it to > sort by ctime instead of the (probably) default mtime. Several other > file download programs set the mtime to the last-modified header or > similar, but they tend not to touch the ctime. > This ended up being the best solution. It seems to work well. Sort of odd in a way but as long as I can sort by the order I downloaded files, I don't care how strange it is. lol Thanks. Dale :-) :-)
Re: [gentoo-user] Youtube-dl and file time stamps.
Dale wrote: > Simon Thelen wrote: >> [2020-07-15 17:30] Dale >>> Howdy, >> Hi, >> >>> I'm not sure what causes this because it doesn't always do this. When I >>> use youtube-dl to download videos, it sometimes uses the current date and >>> time for the time stamp. I like that because I can sort by date and see >>> new videos. On some sites tho it seems to use the time stamp of the file >>> on the server I am downloading from not when it was put on my system. >>> Sometimes I download a video and it may have a time stamp of years ago, >>> decades sometimes. I looked through the help page but can't find a option >>> to tell it to use local time instead of the time from the remote server >>> file. Needless to say, when it does this, I can't tell which videos I >>> recently downloaded since sorting by time stamps is no longer accurate. >>> It's annoying. >>> >>> Has anyone else noticed this behavior? Is there a way to tell it to stop >>> setting it to really old time stamps? Some option that isn't documented >>> maybe. >> You're probably looking for the --no-mtime option. Depending on what >> you're using to sort your local videos you can always just tell it to >> sort by ctime instead of the (probably) default mtime. Several other >> file download programs set the mtime to the last-modified header or >> similar, but they tend not to touch the ctime. >> > > I found that but wasn't sure if that was what I was looking for or > not. I'll give that a try. I added it to the conf file. Now to wait > until this set of videos finishes. > > Thanks much. > > Dale > > :-) :-) This appears to be affected by the other end in some cases. It seems to help in most places but not all. To work around this, I changed the info used to sort the files. Sort of odd at times how it works but anyway. Thanks for the info. It's not a complete fix but it does help. Dale :-) :-)
Re: [gentoo-user] Strange output when restarting network. Long term issue.
Ashley Dixon wrote: > On Wed, Jul 22, 2020 at 09:13:22PM -0500, Dale wrote: >> See how it receives the address from the router and then gives a prompt >> again, then it mounts network file systems where I should type in a >> command but that isn't me.. To get a prompt again, I have to hit the >> return key. Why does it give me a prompt when it isn't done yet? >> Shouldn't it finish completely before returning to a prompt? This is in >> a Konsole, within KDE. That said, I'm pretty sure it does this on a >> console, ctrl alt F*, screen as well. > I can confirm I get this also, using st (suckless/simple terminal). I have a > VPN > script in my net.eno1 script, so the output might differ slightly from > yours, > although the strange behaviour seems the same. > > [super] root@ad-gentoo-main / # etc/init.d/net.eno1 restart > * Stopping openvpn ... [ > ok ] > * Unmounting network filesystems ... [ > ok ] > * Bringing down interface eno1 > * Stopping dhcpcd on eno1 ... > sending signal TERM to pid 13839 > waiting for pid 13839 to exit [ > ok ] > * Bringing up interface eno1 > * dhcp ... > * Running dhcpcd ... > DUID 00:04:aa:aa:aa:aa:aa:aa:aa:aa:aa:aa:d8:cb:8a:c6:64:cd > eno1: IAID 8a:c6:64:cd > eno1: adding address fe80::fd5:dd1a:7c1c:f59c > eno1: rebinding lease of 192.168.0.3 > eno1: carrier lost > eno1: deleting address fe80::fd5:dd1a:7c1c:f59c > eno1: carrier acquired > eno1: IAID 8a:c6:64:cd > eno1: adding address fe80::fd5:dd1a:7c1c:f59c > eno1: soliciting an IPv6 router > eno1: rebinding lease of 192.168.0.3 > eno1: Router Advertisement from fe80::c23e:fff:fe2a:8b8c > eno1: adding address fda8:3e71:2eb6:0:3c02:829f:d71c:a8b9/64 > eno1: adding address 2a02:c7d:25f3:2800:b8e1:f66c:e550:c1b3/64 > eno1: adding route to fda8:3e71:2eb6::/64 > eno1: adding route to 2a02:c7d:25f3:2800::/64 > eno1: requesting DHCPv6 information > eno1: adding default route via fe80::c23e:fff:fe2a:8b8c > forked to background, child pid 14536 [ > ok ] > * received address [ > ok ] > [super] root@ad-gentoo-main / # * Starting openvpn ... [ > ok ] > * WARNING: openvpn has started, but is inactive > * Mounting network filesystems ... [ > ok ] > > [super] root@ad-gentoo-main / # > > I suspect that this is just a silly bug in which a prompt is > displayed > prematurely, and since the shell has already dispatched the $PS1, it > waits > until is received before showing it again, as it thinks you're in > the > process of typing a command. As the "Starting openvpn ..." text is printed > to > stdout, when you press enter, it assumes an empty command has been entered > and > does nothing but display another prompt. > > As a simple demonstration, see what happens when I type a command `echo > Hello` > before pressing enter: > > [super] root@ad-gentoo-main / # etc/init.d/net.eno1 restart > <... TRIM OUTPUT ...> > [super] root@ad-gentoo-main / # * Starting openvpn ... [ > ok ] > * WARNING: openvpn has started, but is inactive > * Mounting network filesystems ... [ > ok ] > echo Hello > Hello > > [super] root@ad-gentoo-main / # > At least it isn't just me. Sometimes I think bugs/bad features pick on me. While annoying, I guess it is harmless. It just seems that as long as it has been doing this, someone would have raised the hood and looked to see what is causing it and how to fix it. Strange that something weird like this has been around so long. Thanks for the reply. At least I know I'm not alone. Dale :-) :-)
Re: [gentoo-user] xorg-server[elogind,-suid] and starting additional Xorg session from running Xorg session
23.07.2020 22:25, Neil Bothwick пишет: > On Thu, 23 Jul 2020 15:15:04 +0300, i.Dark_Templar wrote: > >> With x11-base/xorg-server-1.20.8[elogind,suid] I could just do "sudo -i >> -u another-user DISPLAY= XAUTHORITY= startx $application $app_args -- >> :$nextdisplay" from running X11 session and get myself a separate new >> X11 session running from different user. >> >> With x11-base/xorg-server-1.20.8-r1[elogind,suid] it is also possible to >> do this if line 'allowed_users = anybody' is added to file >> '/etc/X11/X11/Xwrapper.config'. >> >> But with x11-base/xorg-server-1.20.8-r1[elogind,-suid] I couldn't make a >> similar setup to work. I've tried adding options '-keeptty' or 'vt?' or >> both, but all I get are errors like these: >> >> Fatal server error: >> (EE) parse_vt_settings: Cannot open /dev/tty0 (Permission denied) > > Is your new user a member of the tty group? > > No. Should I add every user I wish to allow running Xorg without suid in such setup to tty group? I don't like such idea. Currently, there are no users in this group. Granting a user permissions to control every tty looks like an overkill and an insecure setting. I'm not trying to fix this setup at any cost. I'm trying to figure out if it's possible to do this without suid and I'm just missing something, or if I should stick to suid for my use-case.
Re: [gentoo-user] xorg-server[elogind,-suid] and starting additional Xorg session from running Xorg session
On Thu, 23 Jul 2020 15:15:04 +0300, i.Dark_Templar wrote: > With x11-base/xorg-server-1.20.8[elogind,suid] I could just do "sudo -i > -u another-user DISPLAY= XAUTHORITY= startx $application $app_args -- > :$nextdisplay" from running X11 session and get myself a separate new > X11 session running from different user. > > With x11-base/xorg-server-1.20.8-r1[elogind,suid] it is also possible to > do this if line 'allowed_users = anybody' is added to file > '/etc/X11/X11/Xwrapper.config'. > > But with x11-base/xorg-server-1.20.8-r1[elogind,-suid] I couldn't make a > similar setup to work. I've tried adding options '-keeptty' or 'vt?' or > both, but all I get are errors like these: > > Fatal server error: > (EE) parse_vt_settings: Cannot open /dev/tty0 (Permission denied) Is your new user a member of the tty group? -- Neil Bothwick Old hitchhikers never die-they just throw in the towel. pgpxgmHYN0ArJ.pgp Description: OpenPGP digital signature
Re: [gentoo-user] xorg-server[elogind,-suid] and starting additional Xorg session from running Xorg session
On Thu, 23 Jul 2020 19:38:21 +0300, i.Dark_Templar wrote: > I'm using --newuse (-N). According to 'man emerge', --newuse and > --changed-use are pretty similar, but if disabled USE-flag is added or > removed for package without version change, --changed-use does not > trigger rebuild of package. That's not quite right. The difference is that --changed-use won't trigger a rebuild if the change of use flag makes no difference on your settings whereas --newuse will always rebuild for a changed flag. It has noting to sdo with version changes, which will always case a rebuild because you are using -u. -- Neil Bothwick Adolescence, n.: The stage between puberty and adultery. pgpB_lX8NkfwU.pgp Description: OpenPGP digital signature
Re: [gentoo-user] xorg-server[elogind,-suid] and starting additional Xorg session from running Xorg session
i.Dark_Templar wrote: > 23.07.2020 19:29, Matt Connell (Gmail) пишет: >> On Thu, 2020-07-23 at 19:24 +0300, i.Dark_Templar wrote: >>> 23.07.2020 19:05, Walter Dnes пишет: >>> On Thu, Jul 23, 2020 at 03:15:04PM +0300, i.Dark_Templar wrote > Hi. > I've tried using xorg-server[elogind,-suid] and got an issue. I know this may sound too simple, but did you update world? News item https://www.gentoo.org/support/news-items/2020-06-24-xorg-server-dropping-default-suid.html says... >>> >>> Yes, of course. I usually do 'emerge -avuDN system world' >> I may be way off base, but would the changed-use flag (-U / --changed- >> use) have been needed in order to apply this change? >> >> > I'm using --newuse (-N). According to 'man emerge', --newuse and > --changed-use are pretty similar, but if disabled USE-flag is added or > removed for package without version change, --changed-use does not > trigger rebuild of package. > > Anyway, I just tried running 'emerge -avuUDN system world', and it > reported 'Nothing to merge'. > > Just a FYI. If you put world as a set, you can leave out system. The world set will pull in the system set so it will save you some typing. Just a thought. Dale
Re: [gentoo-user] xorg-server[elogind,-suid] and starting additional Xorg session from running Xorg session
23.07.2020 19:29, Matt Connell (Gmail) пишет: > On Thu, 2020-07-23 at 19:24 +0300, i.Dark_Templar wrote: >> 23.07.2020 19:05, Walter Dnes пишет: >> >>> On Thu, Jul 23, 2020 at 03:15:04PM +0300, i.Dark_Templar wrote Hi. I've tried using xorg-server[elogind,-suid] and got an issue. >>>I know this may sound too simple, but did you update world? News item >>> https://www.gentoo.org/support/news-items/2020-06-24-xorg-server-dropping-default-suid.html >>> says... >> >> >> Yes, of course. I usually do 'emerge -avuDN system world' > > I may be way off base, but would the changed-use flag (-U / --changed- > use) have been needed in order to apply this change? > > I'm using --newuse (-N). According to 'man emerge', --newuse and --changed-use are pretty similar, but if disabled USE-flag is added or removed for package without version change, --changed-use does not trigger rebuild of package. Anyway, I just tried running 'emerge -avuUDN system world', and it reported 'Nothing to merge'.
Re: [gentoo-user] xorg-server[elogind,-suid] and starting additional Xorg session from running Xorg session
On Thu, 2020-07-23 at 19:24 +0300, i.Dark_Templar wrote: > 23.07.2020 19:05, Walter Dnes пишет: > > > On Thu, Jul 23, 2020 at 03:15:04PM +0300, i.Dark_Templar wrote > > > Hi. > > > I've tried using xorg-server[elogind,-suid] and got an issue. > >I know this may sound too simple, but did you update world? News item > > https://www.gentoo.org/support/news-items/2020-06-24-xorg-server-dropping-default-suid.html > > says... > > > Yes, of course. I usually do 'emerge -avuDN system world' I may be way off base, but would the changed-use flag (-U / --changed- use) have been needed in order to apply this change?
Re: [gentoo-user] xorg-server[elogind,-suid] and starting additional Xorg session from running Xorg session
23.07.2020 19:05, Walter Dnes пишет: > On Thu, Jul 23, 2020 at 03:15:04PM +0300, i.Dark_Templar wrote >> Hi. >> >> I've tried using xorg-server[elogind,-suid] and got an issue. > > I know this may sound too simple, but did you update world? News item > https://www.gentoo.org/support/news-items/2020-06-24-xorg-server-dropping-default-suid.html > says... > Yes, of course. I usually do 'emerge -avuDN system world', and I have following options in make.conf: EMERGE_DEFAULT_OPTS="--with-bdeps=y --binpkg-respect-use=y --autounmask=n --complete-graph=y --keep-going" And I did following sequence after emerging xorg-server with different USE-flags before testing it: logout out of X11 session, login into console session as root, restart xdm service (restart sddm), login into KDE session via SDDM. KDE session works fine, but I can't start one more X11 session from it. While I didn't test it, I guess X11 session could be started fine from user if you login into a console session (text-only session), but when I try to start one more X11 session from already running X11 session, I hit insufficient permissions error. >> to globally enable 'elogind' USE flag and update the system >> >> # emerge --newuse @world >> >> Afterwards, one will need to re-login, so the PAM can assign a >> seat. One can confirm that a seat has been assigned upon login >> by running: >> >> $ loginctl user-status > As far as I can see elogind works fine for me in usual scenario: login via SDDM.
Re: [gentoo-user] xorg-server[elogind,-suid] and starting additional Xorg session from running Xorg session
On Thu, Jul 23, 2020 at 03:15:04PM +0300, i.Dark_Templar wrote > Hi. > > I've tried using xorg-server[elogind,-suid] and got an issue. I know this may sound too simple, but did you update world? News item https://www.gentoo.org/support/news-items/2020-06-24-xorg-server-dropping-default-suid.html says... > to globally enable 'elogind' USE flag and update the system > > # emerge --newuse @world > > Afterwards, one will need to re-login, so the PAM can assign a > seat. One can confirm that a seat has been assigned upon login > by running: > > $ loginctl user-status -- Walter Dnes I don't run "desktop environments"; I run useful applications
Re: [gentoo-user] Re: dns/bind-tools 9.14 -> 9.16 pulling in 17 new dependencies?!
On 23/07/20 06:45, Walter Dnes wrote: > ...does indeed pull in sphinx. If I didn't know any better, I'd say > that Lennart is behind this. Anyhow, I've managed to avoid llvm > altogether (USE="-llvm"), so I don't have that problem. I think you might find sphinx is a dependency of the kernel ... Iirc all of the kernel doc is slowly being switched to sphinx. Cheers, Wol
[gentoo-user] xorg-server[elogind,-suid] and starting additional Xorg session from running Xorg session
Hi. I've tried using xorg-server[elogind,-suid] and got an issue. With x11-base/xorg-server-1.20.8[elogind,suid] I could just do "sudo -i -u another-user DISPLAY= XAUTHORITY= startx $application $app_args -- :$nextdisplay" from running X11 session and get myself a separate new X11 session running from different user. With x11-base/xorg-server-1.20.8-r1[elogind,suid] it is also possible to do this if line 'allowed_users = anybody' is added to file '/etc/X11/X11/Xwrapper.config'. But with x11-base/xorg-server-1.20.8-r1[elogind,-suid] I couldn't make a similar setup to work. I've tried adding options '-keeptty' or 'vt?' or both, but all I get are errors like these: Fatal server error: (EE) parse_vt_settings: Cannot open /dev/tty0 (Permission denied) or Fatal server error: (EE) xf86OpenConsole: Cannot open virtual console 5 (Permission denied) Is it possible to make setup like this work with elogind without suid?
Re: [gentoo-user] Re: dns/bind-tools 9.14 -> 9.16 pulling in 17 new dependencies?!
On Thursday, 23 July 2020 06:45:27 BST Walter Dnes wrote: > I assume you're on unstable? Yes, I needed it for zoom. > ACCEPT_KEYWORDS="~amd64" USE="-doc -libffi -ncurses emerge -pv > =sys-devel/llvm-10.0.1 > > ...does indeed pull in sphinx. If I didn't know any better, I'd say > that Lennart is behind this. Anyhow, I've managed to avoid llvm > altogether (USE="-llvm"), so I don't have that problem. I don't have that option, as my video card driver requires llvm. -- Regards, Peter.
Re: [gentoo-user] Simple replacement for "getmail"?
On Wed, Jul 22, 2020 at 12:01:02PM +0200, Matthias Hanft wrote > Walter Dnes wrote: > > > > Would "fetchmail" work as a drop-in replacement for getmail here? Are > > there any better, simpler solutions? > > I don't know "getmail", but "fetchmail" runs here since 10 years > without any problems. Just put a line like > > poll securepop.t-online.de proto pop3 > user "mail@there", with password "pass@there", is "mail@here" here, > ssl; > > into /etc/fetchmailrc, adjust polling_period="300" in > /etc/conf.d/fetchmail, and that's it. (Of course, the usual > stuff like "/etc/init.d/fetchmail start" and "rc-update > add fetchmail default".) I'd prefer to launch manually as required. According to weboage https://calomel.org/fetchmailrc.html I can set up $HOME/.fetchmailrc I want to pull email from my local ISP and from 3 other (remote) servers. Notes: 1) I want to run in "single-drop mode". 2) I do *NOT* want bouncing under any circumstances. 3) My procmail script (built up over the years) decides which local inbox an individual email goes to. Here's a first attempt at a ~/.fetchmailrc. I'll have to insert correct names later... set logfile fetchmail.log poll inmail.myisp.com proto POP3 user "remote_username" pass "PASSWORD=" is "local_username" preconnect "date >> fetchmail.log" #ssl fetchall no keep no rewrite mda "/usr/bin/procmail -f %F"; set logfile fetchmail.log poll inmail.server2.com proto POP3 user "remote_username" pass "PASSWORD=" is "local_username" preconnect "date >> fetchmail.log" ssl fetchall no keep no rewrite mda "/usr/bin/procmail -f %F"; set logfile fetchmail.log poll inmail.server3.com proto POP3 user "remote_username" pass "PASSWORD=" is "local_username" preconnect "date >> fetchmail.log" ssl fetchall no keep no rewrite mda "/usr/bin/procmail -f %F"; set logfile fetchmail.log poll inmail.server4.com proto POP3 user "remote_username" pass "PASSWORD=" is "local_username" preconnect "date >> fetchmail.log" ssl fetchall no keep no rewrite mda "/usr/bin/procmail -f %F"; -- Walter Dnes I don't run "desktop environments"; I run useful applications
Re: [gentoo-user] Re: dns/bind-tools 9.14 -> 9.16 pulling in 17 new dependencies?!
On Thu, Jul 23, 2020 at 01:45:27AM -0400, Walter Dnes wrote: > ACCEPT_KEYWORDS="~amd64" USE="-doc -libffi -ncurses emerge -pv > =sys-devel/llvm-10.0.1 > > ...does indeed pull in sphinx. If I didn't know any better, I'd say > that Lennart is behind this. Anyhow, I've managed to avoid llvm > altogether (USE="-llvm"), so I don't have that problem. As in most other Sphinx-utilising packages, it is listed in the BDEPEND variable, meaning that it is only required during the building of the package [1]. If `emerge -c` isn't removing what you would expect, try something like `emerge -cv dev-python/sphinx` to see the explicit reverse-dependencies blocking the removal. [1] https://devmanual.gentoo.org/general-concepts/dependencies/index.html#build-dependencies -- Ashley Dixon suugaku.co.uk 2A9A 4117 DA96 D18A 8A7B B0D2 A30E BF25 F290 A8AA signature.asc Description: PGP signature
