Re: [gentoo-user] Re: dev-ruby/json-1.8.0
This is what I have. stephen #grep RUBY /etc/portage/make.conf RUBY_TARGETS=ruby20 stephen # ls -l /usr/bin/rdoc lrwxrwxrwx 1 root root 6 Jun 6 20:13 /usr/bin/rdoc - rdoc20 stephen # eselect ruby list Available Ruby profiles: [1] ruby19 (with Rubygems) [2] ruby20 (with Rubygems) * Regards On Sun, Jun 8, 2014 at 2:20 AM, walt w41...@gmail.com wrote: On 06/07/2014 12:56 AM, Hans de Graaff wrote: On Fri, 06 Jun 2014 15:47:38 -0700, walt wrote: Is all of the above familiar to you? If not, you may need more help with managing multiple ruby versions. I find it a large PITA and I could use more help myself :) Could you explain what bothers you or where you would need help? Hi Hans. The annoying problems occur when updating ruby-related packages. For example, I (want to) use only ruby19: #grep RUBY /etc/portage/make.conf RUBY_TARGETS=ruby19 In spite of that, portage often insists on installing other versions of ruby, rdoc, rubygems, and you already know the others. AFAICT, the other versions of ruby are dragged in by old ruby packages that were installed before I started using RUBY_TARGETS (because I didn't yet know about RUBY_TARGETS), I discovered all of this by grepping for ruby in /var/db/pkg but it took me a long time to get it sorted out, and I don't expect that a gentoo beginner could do it. (OTOH maybe a gentoo beginner wouldn't care about installing multiple ruby versions :) Thanks for taking the time to read gentoo.user and even more thanks for being a gentoo dev :)
[gentoo-user] Zsh completion
Hi I use zsh and have quite perfect completion setup with it. There is just one very annoying feature that I have failed to switch off. With paths when I type this: cd /archives/NE tab zsh produces: cd /achieves2/NEW/ The archives directory does not contain NEW directory and archives2 does. I would want that zsh wouldn't modify anything but the current level path I'm writing. So in this case it should of shown empty cuz there are no options to choose from. This happens of course with any similar directory case. The annoyance is that I know where I'm going and the right NEW directory in this case is under /archives/movies/NEW and not the one under /archives2/. So I have to clear some of the text which is slow :( Would anyone know how to correct this¿? I have tried various options of approximation... Actually I don't like the approximation at all and have tried to fully disable it... -- Matti
[gentoo-user] Re: dev-ruby/json-1.8.0
On Sat, 07 Jun 2014 17:20:22 -0700, walt wrote: On 06/07/2014 12:56 AM, Hans de Graaff wrote: For example, I (want to) use only ruby19: #grep RUBY /etc/portage/make.conf RUBY_TARGETS=ruby19 Yes, in hindsight I think that should have been the current default since ruby19 has the best overall coverage for packages. Once ruby20 has caught up I think we'll move to a default of RUBY_TARGETS=ruby20 In spite of that, portage often insists on installing other versions of ruby, rdoc, rubygems, and you already know the others. Partially this was because we tried to solve another issue when ruby20 went stable. I removed those forced use flags for ruby20 last week, so this should no longer happen. We still need to come up with a good plan when the same issue will pop up for ruby21. AFAICT, the other versions of ruby are dragged in by old ruby packages that were installed before I started using RUBY_TARGETS (because I didn't yet know about RUBY_TARGETS), Yes, these will still have other ruby targets recorded and thus also request them for their dependencies. emerge --newuse should be able to help here. I discovered all of this by grepping for ruby in /var/db/pkg but it took me a long time to get it sorted out, and I don't expect that a gentoo beginner could do it. (OTOH maybe a gentoo beginner wouldn't care about installing multiple ruby versions :) We try to keep the default settings so that someone who doesn't care or know about ruby should get a good experience. Moving from ruby18 to ruby19 we did some things that could have been handled better (such as not mentioning that the new ruby must be eselected before making the switch), so hopefully we've learned from those when we do the next update. Hans
[gentoo-user] Ifplugd breaks services
Hi I also have other problems in my life. One of them is on one of my gentoo server. This server has two network cards one serves intranet and the other internet. The on that is on the internet is attached to a cable modem. The modem is buggy and some times reboots it self losing the link so I have ifplugd there get new address via dhcp immediately. Intranet card is configured not to use ifplugd. I'm using OpenRC. The problems are related to iptables and samba. Samba: when ifplugd runs down the internet card samba is killed. This shouldn't happen. Samba is configured only to use intranet card. Samba always fails to start when ifplugd starts the internet card. Manual starting is required. Iptables: the system uses new nic names (enp7s0 etc). Iptables has them correctly in the rules and in rules save. However when ifplugd cycles the internet nic all the nic names in the in-kernel rules change to eth0 an eth1. I need to zap iptables and then start it to reset the rules. Any suggestions where to start? Or just disable ifplugd? -- Matti
Re: [gentoo-user] Re: dev-ruby/json-1.8.0
Okay I am now using ruby19, This have solved my problem. Thanks stephen # eselect ruby list Available Ruby profiles: [1] ruby19 (with Rubygems) * [2] ruby20 (with Rubygems) stephen # ls -l /usr/bin/rdoclrwxrwxrwx 1 root root 6 Jun 8 11:45 /usr/bin/rdoc - rdoc19 stephen # grep RUBY /etc/portage/make.conf RUBY_TARGETS=ruby19 On Sun, Jun 8, 2014 at 10:39 AM, Hans de Graaff gra...@gentoo.org wrote: On Sat, 07 Jun 2014 17:20:22 -0700, walt wrote: On 06/07/2014 12:56 AM, Hans de Graaff wrote: For example, I (want to) use only ruby19: #grep RUBY /etc/portage/make.conf RUBY_TARGETS=ruby19 Yes, in hindsight I think that should have been the current default since ruby19 has the best overall coverage for packages. Once ruby20 has caught up I think we'll move to a default of RUBY_TARGETS=ruby20 In spite of that, portage often insists on installing other versions of ruby, rdoc, rubygems, and you already know the others. Partially this was because we tried to solve another issue when ruby20 went stable. I removed those forced use flags for ruby20 last week, so this should no longer happen. We still need to come up with a good plan when the same issue will pop up for ruby21. AFAICT, the other versions of ruby are dragged in by old ruby packages that were installed before I started using RUBY_TARGETS (because I didn't yet know about RUBY_TARGETS), Yes, these will still have other ruby targets recorded and thus also request them for their dependencies. emerge --newuse should be able to help here. I discovered all of this by grepping for ruby in /var/db/pkg but it took me a long time to get it sorted out, and I don't expect that a gentoo beginner could do it. (OTOH maybe a gentoo beginner wouldn't care about installing multiple ruby versions :) We try to keep the default settings so that someone who doesn't care or know about ruby should get a good experience. Moving from ruby18 to ruby19 we did some things that could have been handled better (such as not mentioning that the new ruby must be eselected before making the switch), so hopefully we've learned from those when we do the next update. Hans
Re: [gentoo-user] Re: Logitech C920 HD Pro Webcamera ... view: YES record: NO ?
James wirel...@tampabay.rr.com [14-06-07 17:52]: meino.cramer at gmx.de writes: the Logitech c920 HD Pro webcam is able to deliver 1920xq080x30fps (H.264). Some interesting notes and comments at the bottom of this page: http://www.ideasonboard.org/uvc/ Also, in the past, I have found the best comments on various camera hardware in the relevant kernel driver. Sometimes that is a unique driver, other times it is a unified kernel driver for a similar group of video cameras. Sometimes you can go way back to where the kernel driver for a given video camera first appeared (got supported) in the linux kernel for detailed information on the limitations of a given piece of hardware. Often, the poorer performance on Linux, was intentionally due to the actions of the manufacturer, particular on max frame rate, bit minipulations and other key parameter settings of the cameras. Once folks learned the protocols (decoded them) on windows, they would make those adjustments with windows software and reverse engineer the driver software so as to be able to support various linux drivers. Logitech is very reasonable (at least compared to other vendors) but even there newer hardware rarely works with the best of a given feature set, until the product has been out for a while. If not, Logitech should have a published interface specification so and to make reverse engineering not necessary? Dunno know the specifics on the model your listing, as I've been out of that 'game' for a while now, but all those tigers still have the same stripes Good hunting! hth, James Hi, I've git somethiong working for me: With vlc Capture device and a lot of setting I am able to wath and record if wanted the stream of the c920. BUT There is big delay between audio and video (audio delayed) and a smaller delay between reality and strem (which doesn't matter in my case). How can I fix the delay between video and audio? Best regards, mcc
Re: [gentoo-user] Ifplugd breaks services
On Sunday 08 Jun 2014 10:25:40 Matti Nykyri wrote: Hi I also have other problems in my life. One of them is on one of my gentoo server. This server has two network cards one serves intranet and the other internet. The on that is on the internet is attached to a cable modem. The modem is buggy and some times reboots it self losing the link so I have ifplugd there get new address via dhcp immediately. Intranet card is configured not to use ifplugd. I'm using OpenRC. Are you sure of this? How have you configured your intranet card to not be acted upon by ifplugd? From what I see, ifplugd will pick up any interface in /etc/init.d: EXEC=/etc/init.d/net.$1 The problems are related to iptables and samba. Samba: when ifplugd runs down the internet card samba is killed. This shouldn't happen. Samba is configured only to use intranet card. Samba always fails to start when ifplugd starts the internet card. Manual starting is required. Iptables: the system uses new nic names (enp7s0 etc). Iptables has them correctly in the rules and in rules save. However when ifplugd cycles the internet nic all the nic names in the in-kernel rules change to eth0 an eth1. I need to zap iptables and then start it to reset the rules. This does not happen here. When ifplugd restarts a NIC it always comes back with the new consistent naming. Do you have some udev rules defined which are picked up on the second time that the ifplugd brings up the card, but not the first? Any suggestions where to start? Or just disable ifplugd? Have you tried configuring static IP addresses, broadcast and gateways for each card using the /etc/cong.d/net file? You can still set dhcp as a fall back if you want, although I am not sure it is necessary for a server in a static LAN. The file /usr/share/doc/netifrc-0.2.2/net.example.bz2 provides suggestions for different set ups. -- Regards, Mick signature.asc Description: This is a digitally signed message part.
Re: [gentoo-user] Ifplugd breaks services
On Sun, Jun 08, 2014 at 11:25:53AM +0100, Mick wrote: On Sunday 08 Jun 2014 10:25:40 Matti Nykyri wrote: Hi I also have other problems in my life. One of them is on one of my gentoo server. This server has two network cards one serves intranet and the other internet. The on that is on the internet is attached to a cable modem. The modem is buggy and some times reboots it self losing the link so I have ifplugd there get new address via dhcp immediately. Intranet card is configured not to use ifplugd. I'm using OpenRC. Are you sure of this? How have you configured your intranet card to not be acted upon by ifplugd? From what I see, ifplugd will pick up any interface in /etc/init.d: EXEC=/etc/init.d/net.$1 Actually it's not ifplugd's fault. It is just the one that restarts services... The restarting is the thing that breaks stuff: server% [13:44] /var/log$ sudo iptables -v -L -t nat Chain POSTROUTING (policy ACCEPT 10142 packets, 743K bytes) pkts bytes target prot opt in out source destination 8307 616K MASQUERADE all -- anyenp0s10 anywhere anywhere server% [13:45] /var/log$ sudo /etc/init.d/net.enp0s10 stop * Stopping NIS Server ... [ ok ] * samba - stop: smbd ... [ ok ] * samba - stop: nmbd ... * start-stop-daemon: 2 process(es) refused to stop [ !! ] * Unmounting network filesystems ...[ ok ] * Stopping chrooted named ... * Umounting chroot dirs ... * umounting /chroot/dns/usr/share/GeoIP ... [ ok ] * umounting /chroot/dns/etc/bind ...[ ok ] * umounting /chroot/dns/var/log/named ... [ ok ] * umounting /chroot/dns/var/bind ...[ ok ] * Stopping dhcpd ...[ ok ] * Bringing down interface enp0s10 * Stopping dhclient on enp0s10 ...[ ok ] * Stopping ifplugd on enp0s10 ... [ ok ] server% [13:45] /var/log$ sudo iptables -v -L -t nat Chain POSTROUTING (policy ACCEPT 10147 packets, 743K bytes) pkts bytes target prot opt in out source destination 8309 617K MASQUERADE all -- anyenp0s10 anywhere anywhere server% [13:45] /var/log$ sudo /etc/init.d/net.enp0s10 start * Bringing up interface enp0s10 * Changing MAC address of enp0s10 ... [ ok ] * changed to 00:80:23:7A:8A:A4 * Starting ifplugd on enp0s10 ... [ ok ] * Backgrounding ... * WARNING: net.enp0s10 has started, but is inactive server% [13:45] /var/log$ sudo iptables -v -L -t nat Chain POSTROUTING (policy ACCEPT 10147 packets, 743K bytes) pkts bytes target prot opt in out source destination 8309 617K MASQUERADE all -- anyenp0s10 anywhere anywhere It takes around 40 seconds for dhclient to address from ISP (net-misc/dhcp-4.2.5_p1) After it gets the address iptables is changed: server% [13:45] /var/log$ sudo iptables -v -L -t nat Chain POSTROUTING (policy ACCEPT 2 packets, 152 bytes) pkts bytes target prot opt in out source destination 0 0 MASQUERADE all -- anyeth1anywhere anywhere server% [13:48] /var/log$ ps aux | grep dhclient root 22011 0.0 0.2 16200 7108 ?Ss 13:46 0:00 /sbin/dhclient -e PEER_NTP=no -e IF_METRIC=3 -q -1 -pf /var/run/dhclient-enp0s10.pid enp0s10 server% [13:48] /var/log$ ls /etc/init.d/net* lrwxrwxrwx 1 root root 6 Oct 4 2011 /etc/init.d/net.enp0s10 - net.lo* lrwxrwxrwx 1 root root 6 Oct 4 2011 /etc/init.d/net.enp5s12 - net.lo* -rwxr-xr-x 1 root root 17412 Jan 2 23:42 /etc/init.d/net.lo* The problems are related to iptables and samba. Samba: when ifplugd runs down the internet card samba is killed. This shouldn't happen. Samba is configured only to use intranet card. Samba always fails to start when ifplugd starts the internet card. Manual starting is required. Iptables: the system uses new nic names (enp7s0 etc). Iptables has them correctly in the rules and in rules save. However when ifplugd cycles the internet nic all the nic names in the in-kernel rules change to eth0 an eth1. I need to zap iptables and then start it to reset the rules. This does not happen here. When ifplugd restarts a NIC it always comes back with the new consistent naming. Do you have some udev rules defined which are picked up on the second time that the ifplugd brings up the card, but not the first? No I don't. And as stated when dhclient sets
[gentoo-user] x11-misc/synergy
Gentooers, Anyone built a multi head setup using synergy? If so, do you like it? Pros and Cons? curiously, James
Re: [gentoo-user] x11-misc/synergy
2014-06-08 8:25 GMT-06:00 James wirel...@tampabay.rr.com: Gentooers, Anyone built a multi head setup using synergy? For me now is essential to have synergy on my laptop and desktop it improves the workflow greatly. If so, do you like it? A lot, and recomend building it with the qt interface it makes simple reorganizing the screens if you move them, or change the computer acting as server. Pros and Cons? Its easy to use, doesn't get in your way, and maybe two cons,it ask you to pay if you want the shared clipboard feature, only one computer acts like server so the control its : laptop - desktop or desktop - laptop not desktop -laptop curiously, James
[gentoo-user] What's with foomatic-filters and cups-filters?
Hi, Gentoo. I'm getting back to trying to update my system again, after having lost the thread back in February. I've lost hour after hour after hour with portage's failure to maintain consistency in its internal structures on my system. Sometimes I think it would have been better for me to have just given up, bought a new PC and installed some other distribution on it. Anyhow, after a recommendation from Sebastian Luther at Gentoo, I ran emerge with debugging enabled, thusly: emerge -p --backtrack=100 --debug icu emerge-debug.log . It failed, of course, as usual. But in the middle of the debugging output (which is 147k lines long), appeared this: [ebuild U ] app-text/ghostscript-gpl-9.10-r2 [9.05-r1] LINGUAS=-de% [ebuild r U ] net-print/cups-filters-1.0.53 [1.0.36-r1] USE=dbus%* foomatic%* [blocks b ] net-print/cups-filters-1.0.36-r2 (net-print/cups-filters-1.0.36-r2 is blocking app-text/ghostscript-gpl-9.10-r2) [ebuild r U ] app-text/evince-3.10.3 [3.8.3] USE=-libsecret% [blocks B ] net-print/foomatic-filters (net-print/foomatic-filters is blocking net-print/cups-filters-1.0.53) [blocks B ] =net-print/cups-filters-1.0.43-r1[foomatic] (=net-print/cups-filters-1.0.43-r1[foomatic] is blocking net-print/foomatic-filters-4.0.17-r1) * Error: The above package list contains packages which cannot be * installed at the same time on the same system. (net-print/cups-filters-1.0.53::gentoo, ebuild scheduled for merge) pulled in by net-print/cups-filters:0 required by @__auto_slot_operator_replace_installed__ =net-print/cups-filters-1.0.43-r1[foomatic] required by (net-print/foomatic-filters-ppds-20070501::gentoo, installed) =net-print/cups-filters-1.0.30 required by (net-print/cups-1.6.4::gentoo, installed) net-print/cups-filters required by (net-print/foomatic-filters-4.0.17-r1::gentoo, installed) (net-print/foomatic-filters-4.0.17-r1::gentoo, installed) pulled in by net-print/foomatic-filters required by @selected net-print/foomatic-filters required by (net-print/cups-1.6.4::gentoo, installed) . What is all this trying to tell me? I've tried for over an hour to make sense of it, but my eyes just glaze over. My best guess is that cups-filters and foomatic-filters are packages which can't be installed together. But I _need_ foomatic-filters - otherwise my printer doesn't print. Or do I? cups-filters seems to be needed by cups. What _are_ cups-filters and foomatic-filters? emerge -s is little help here. Why do I need both of them? How do I disentangle my system? Help, please! -- Alan Mackenzie (Nuremberg, Germany).
Re: [gentoo-user] What's with foomatic-filters and cups-filters?
On Sun, 8 Jun 2014 15:48:09 +, Alan Mackenzie wrote: (net-print/cups-filters-1.0.53::gentoo, ebuild scheduled for merge) pulled in by net-print/cups-filters:0 required by @__auto_slot_operator_replace_installed__ =net-print/cups-filters-1.0.43-r1[foomatic] required by (net-print/foomatic-filters-ppds-20070501::gentoo, installed) =net-print/cups-filters-1.0.30 required by (net-print/cups-1.6.4::gentoo, installed) net-print/cups-filters required by (net-print/foomatic-filters-4.0.17-r1::gentoo, installed) Does the emerge output somewhere include an update to cups? If the installed and new versions have different dependencies you may see this. Remove cups-filters and foomatic-filters, update cups and then run the world update. -- Neil Bothwick Every time I jump on the bandwagon all its wheels fall off. signature.asc Description: PGP signature
Re: [gentoo-user] Ifplugd breaks services
On Sun, 8 Jun 2014 15:01:02 +0300, Matti Nykyri wrote: Actually it's not ifplugd's fault. It is just the one that restarts services... The restarting is the thing that breaks stuff: Are you running ifplugd directly or letting openrc deal with this? The latter is the recommended way for openrc, leave ifplugd installed but don't add it to a runlevel. Does the problem persist if you do this? Have you tried changing rc_depend_strict in /etc/rc.conf? flamebaitOr you could switch to systemd which I suspect could be made to handle this situation better./flamebait :) -- Neil Bothwick I work with User-Surly Software. signature.asc Description: PGP signature
Re: [gentoo-user] What's with foomatic-filters and cups-filters?
Am Sonntag, 8. Juni 2014, 17:48:09 schrieb Alan Mackenzie: . What is all this trying to tell me? I've tried for over an hour to make sense of it, but my eyes just glaze over. My best guess is that cups-filters and foomatic-filters are packages which can't be installed together. But I _need_ foomatic-filters - otherwise my printer doesn't print. Or do I? cups-filters seems to be needed by cups. What _are_ cups-filters and foomatic-filters? emerge -s is little help here. Why do I need both of them? * cups-filters is a former part of cups that provides file format conversions (among other things). Basically it (also) makes sure that everything is internally converted to PDF. It's not part of CUPS (as maintained by Apple) anymore, but hard-required by CUPS on Linux (and maintained by the Linux Foundation). * foomatic-filters is a set of printer drivers, basically. * Some time ago the cups-filters maintainers took over maintainership of the foomatic-filters part for CUPS as well, and integrated it cleanly into cups- filters. That's the reason for the blocker; recent cups-filters contain the newest foomatic code available. The former separate foomatic-filters package is now unmaintained. So, we have the following possibilities for installation: 1) normal CUPS user, recommended, this is what comes by default (unless you do something stupid such as USE=-*) net-print/cups net-print/cups-filters[foomatic] 2) NOT recommended, dead code, unmaintained: net-print/cups net-print/cups-filters[-foomatic] net-print/foomatic-filters 3) for the stone age people out there, NOT recommended, dead code, unmaintained: any other printing system, e.g. lprng net-print/foomatic-filters So, what's wrong in your case? No idea, but after longish not-updating things do get hard for emerge to unravel. My recommendation is, since foomatic- filters and cups-filters are only needed for printing and emerge runs fine without them, force-remove both and let emerge figure out the right package set from scratch. [This basically works with any blocker as a last resort, but can be *very* dangerous for packages that are needed by the core system. You definitely don't want to remove gcc or glibc this way, for example. :)] emerge -aC net-print/cups-filters net-print/foomatic-filters emerge -uDNavt --backtrack=100 world Cheers, Andreas -- Andreas K. Huettel Gentoo Linux developer (council, kde) dilfri...@gentoo.org http://www.akhuettel.de/
Apologies - WAS: Re: [gentoo-user] Systemd upower
On 6/4/2014 9:47 AM, Neil Bothwick n...@digimed.co.uk wrote: You seem to think the Upower devs simply decided to use systemd instead of doing it themselves. In fact, they were always using code, from either systemd or pm-utils. The fact that development stopped on pm-utils is neither the fault of the Upower or systemd people. They were reduced to a choice of one and you blame them for making the wrong choice? Actually, I wasn't talking about upower specifically, I was talking about this whole slippery slope that is systemd - but you are right, and I absolutely apologize for my comment about 'lazy devs', and most of my other negative comments. I still don't like the way systemd seems to be devouring everything to the point that it is apparently inevitable that it will become the default init system for all linux system. But I also admit that this is more just personal bias against Lennart/Kay/etc and all things related to them, all coming just from the many threads I've read, and also just fear of change in general (being that I am *not* a programmer, and am *not* capable of doing anything about this myself, regardless of if I would have the time or not). So, I will absolutely cease and desist denigrating systemd, at least until such time as I can speak from direct personal experience. First question: is there a decent guide to installing a gentoo system from scratch using systemd as the init system? Second question: is there a decent guide to how to switch from OpenRC to systemd? Third question: is there a decent guide on how to switch from systemd back to OpenRC, if I encounter any serious problems on a production box? Thanks, and again, my apologies for starting another flame-fest, and especially for basically abandoning the thread afterwards (busy week last week)...
Re: [gentoo-user] Ifplugd breaks services
On Jun 8, 2014, at 19:15, Neil Bothwick n...@digimed.co.uk wrote: On Sun, 8 Jun 2014 15:01:02 +0300, Matti Nykyri wrote: Actually it's not ifplugd's fault. It is just the one that restarts services... The restarting is the thing that breaks stuff: Are you running ifplugd directly or letting openrc deal with this? The latter is the recommended way for openrc, leave ifplugd installed but don't add it to a runlevel. Does the problem persist if you do this? Ifplugd package doesn't have anything installed in init.d/ so it's not added to any runlevel. Have you tried changing rc_depend_strict in /etc/rc.conf? Setting rc_depend_strict to NO, fixes the problem :) With that set to YES all the services are killed. So I'll stick with NO. Still I think that all services stopped should be restarted by default. flamebaitOr you could switch to systemd which I suspect could be made to handle this situation better./flamebait :) I rather not ;) -- -Matti
Re: [gentoo-user] Ifplugd breaks services
On Sun, 8 Jun 2014 20:44:47 +0300, Matti Nykyri wrote: Have you tried changing rc_depend_strict in /etc/rc.conf? Setting rc_depend_strict to NO, fixes the problem :) With that set to YES all the services are killed. So I'll stick with NO. Still I think that all services stopped should be restarted by default. Yes, it does seem like a bug, or at least an undocumented feature. flamebaitOr you could switch to systemd which I suspect could be made to handle this situation better./flamebait :) I rather not ;) You're already using some Lennartware so you're already on the slippery slope :-O -- Neil Bothwick WinErr 678: This will end your Windows session. Do you want to play another game? signature.asc Description: PGP signature
[gentoo-user] udev 208 to 212 update, 2 questions...
Ok, Getting ready to do this update, but the wiki text is confusing... It states: udev 208 to 212 The following special attention is required: snip File /lib/udev/rules.d/80-net-name-slot.rules was replaced with /lib/udev/rules.d/80-net-setup-link.rules. If you are currently using an empty (or single-comment) /etc/udev/rules.d/80-net-name-slot.rules to disable predictable network interface names, you should now use 80-net-setup-link.rules. eg: cd /etc/udev/rules.d/ ln 80-net-name-slot.rules 80-net-setup-link.rules ..to keep the override both pre- and post-upgrade; you can then: rm /etc/udev/rules.d/80-net-name-slot.rules ..once you have upgraded. The hardlink can be made now, in order to protect against not noticing the upgrade in a busy or non-professional situation. However, 80-net-setup-link.rules is only a trigger for the actual configuration file 99-default.link at /lib/systemd/network/ which you can override at /etc/systemd/network/ The most reliable way of disabling the new network interface scheme is still the kernel bootline parameter: net.ifnames=0 * Question 1: What if I am not using an empty (or single comment) /etc/udev/rules.d/80-net-name-slot.rules? Does this mean I can ignore everything that follows (the comment really should open with that if so)? Question 2: What is up with the last section talking about the net setup rules only with respect to systemd? OpenRC is currently still the default init system for gentoo if I'm not mistaken, so why does this comment only reference systemd, totally ignoring OpenRC users? Thanks, Charles
[gentoo-user] Re: Only 4 of 8 GB usable
Alex Schuster wo...@wonkology.org schrieb: So I installed another 4 GiB RAM into a Gentoo amd64 system that had 4 GiB already. But it still sees only 4 GiB, not 8 GiB: leela ~ # uname -a Linux leela 3.6.11-gentoo #3 SMP Mon Feb 4 15:37:48 CET 2013 x86_64 AMD A6-3500 APU with Radeon(tm) HD Graphics AuthenticAMD GNU/Linux leela ~ # free -m total used free sharedbuffers cached Mem: 3688 3269419 0108 1050 -/+ buffers/cache: 2110 1577 Swap: 2047 54 1993 Huh? Any idea why this is? The BIOS shows the full 8GiB, and lshw finds it. dmidecode shows that 8G should work: Two ideas: An addon card is blocking the usage of RAM above 4 GB. See if your BIOS is enabled for 64 bit systems and if everything will map above 32 bit address space. Usually there are BIOS configuration sections like resource allocation, memory hole, resource relocation, or similar - you get the idea. My second idea is: Look at dmesg if there are errors from MTRR which allow the kernel to only make use of 4 GB of RAM. Usually, the kernel tries its best to adjust MTRR to make as much use of RAM as possible with lower priority to memory access performance. If you tuned MTRR manually before, you are likely to be running into memory issues. Also, check if there is a BIOS update available. Sometimes this is needed to correct MTRR settings so memory becomes avaiable to the OS. Also look at your kernel options regarding MTRR support / defaults. These kernel options may help you: mtrr_gran_size, mtrr_chunk_size E.g., I have 16 GB installed and adjusted MTRR settings to allow using write combining (which cannot be enabled by the kernel with the default MTRR settings from BIOS). But that came with the downside of loosing around 1 GB of usable RAM. -- Replies to list only preferred.
Re: [gentoo-user] udev 208 to 212 update, 2 questions...
On Sun, Jun 8, 2014 at 2:34 PM, Tanstaafl tansta...@libertytrek.org wrote: Question 1: What if I am not using an empty (or single comment) /etc/udev/rules.d/80-net-name-slot.rules? Does this mean I can ignore everything that follows (the comment really should open with that if so)? Yes Question 2: What is up with the last section talking about the net setup rules only with respect to systemd? OpenRC is currently still the default init system for gentoo if I'm not mistaken, so why does this comment only reference systemd, totally ignoring OpenRC users? Nothing in this news item has anything to do with anybody who is using the sys-apps/systemd package (at least, they're not impacted by this particular upgrade). sys-fs/udev installs its default network configuration rules in the path /lib/systemd/network/, and if you want to override them you do this in /etc/systemd/network. If you do have sys-apps/systemd installed then you can override the rules in the same place, but the news item is about an upgrade to sys-fs/udev. If you missed the news, systemd took over udev ages ago, and has been renaming paths/binaries/etc. It still does the same things it did before the takeover. You will just see more references to systemd in path names. With just udev the only thing the network scripts do is control interface naming (I think). If you do migrate to systemd the more recent versions include systemd-networkd which can also configure the network. I've found that this works fairly well so far. It seems to be keeping my dhcp lease alive, which is more than I can say about dhcpcd after the last dracut upgrade (the box would drop off the network every 24h, and then lock up being that root was via nfs). Rich
Re: Apologies - WAS: Re: [gentoo-user] Systemd upower
On Sun, Jun 8, 2014 at 1:26 PM, Tanstaafl tansta...@libertytrek.org wrote: First question: is there a decent guide to installing a gentoo system from scratch using systemd as the init system? I've done this a few times on VMs. Just follow the handbook, but skip steps about configuring hostname/timezone/locale/etc since systemd does this (but do set up locale.gen). Then follow the systemd install guide. If you follow both guides to completion you won't hurt anything, but you'll just end up configuring some things twice (but systemd does migrate some of your settings over). Second question: is there a decent guide to how to switch from OpenRC to systemd? Yes, the systemd wiki page is the best place to go for this. It is pretty straightfoward. The only thing I'd do differently is just use networkd. The guide doesn't include that yet. cat /etc/systemd/network/dhcp.network [Match] Name=en* [Network] DHCP=yes --- end file --- (as long as you keep the extension you can call that file whatever you want, and if your interface doesn't match that glob you can tweak it) Also, if you have any network filesystems be sure to set the _netdev option in fstab. Third question: is there a decent guide on how to switch from systemd back to OpenRC, if I encounter any serious problems on a production box? For the most part you can just change the init setting on your kernel line to switch back and forth. You'll end up using udev packaged with systemd, but for the most part that shouldn't cause too many problems. Oh, if you're using dracut there is a chance it won't realize you aren't running systemd in your kernel and that could cause some issues (I was getting some of that before I intended to cut over to systemd in my last migration, but I didn't mess with it for long). Just keep in mind that immediately following the migration you won't have any services enabled. That means no network, no sshd, etc. Starting that stuff up is pretty easy, but it is just like having a fresh OpenRC install. Rich
Re: [gentoo-user] x11-misc/synergy
Can't live without it. I use it at work and at home. I actually paid for the premium for the shared clipboard and to keep the program going. On Jun 8, 2014 11:13 AM, Jc García jyo.gar...@gmail.com wrote: 2014-06-08 8:25 GMT-06:00 James wirel...@tampabay.rr.com: Gentooers, Anyone built a multi head setup using synergy? For me now is essential to have synergy on my laptop and desktop it improves the workflow greatly. If so, do you like it? A lot, and recomend building it with the qt interface it makes simple reorganizing the screens if you move them, or change the computer acting as server. Pros and Cons? Its easy to use, doesn't get in your way, and maybe two cons,it ask you to pay if you want the shared clipboard feature, only one computer acts like server so the control its : laptop - desktop or desktop - laptop not desktop -laptop curiously, James
Re: [gentoo-user] What's with foomatic-filters and cups-filters?
Hi, Neil. On Sun, Jun 08, 2014 at 05:11:20PM +0100, Neil Bothwick wrote: On Sun, 8 Jun 2014 15:48:09 +, Alan Mackenzie wrote: (net-print/cups-filters-1.0.53::gentoo, ebuild scheduled for merge) pulled in by net-print/cups-filters:0 required by @__auto_slot_operator_replace_installed__ =net-print/cups-filters-1.0.43-r1[foomatic] required by (net-print/foomatic-filters-ppds-20070501::gentoo, installed) =net-print/cups-filters-1.0.30 required by (net-print/cups-1.6.4::gentoo, installed) net-print/cups-filters required by (net-print/foomatic-filters-4.0.17-r1::gentoo, installed) Does the emerge output somewhere include an update to cups? If the installed and new versions have different dependencies you may see this. Yes, I think it did. Remove cups-filters and foomatic-filters, update cups and then run the world update. Well, the first bit of that went well indeed. Updating cups pulled in cups-filters but not foomatic-filters. Printing now seems to work without the latter. As for a world update, I'm still summoning up the courage. I might let that run overnight, after doing a backup. Thanks for the help! -- Neil Bothwick -- Alan Mackenzie (Nuremberg, Germany).
[gentoo-user] Re: Logitech C920 HD Pro Webcamera ... view: YES record: NO ?
meino.cramer at gmx.de writes: I've git somethiong working for me: With vlc Capture device and a lot of setting I am able to wath and record if wanted the stream of the c920. BUT There is big delay between audio and video (audio delayed) and a smaller delay between reality and strem (which doesn't matter in my case). How can I fix the delay between video and audio? http://www.videolan.org/developers/x264.html Is my recommendation to ask the folks for some syntax snippets. I use to have a link to all sorts of command line examples to manually filter, convert and perform all sorts of video magic. I cannot find the link now. These x.264 devs a VLC/videolan are the best open source h.264 coders around good hunting! James
Re: [gentoo-user] What's with foomatic-filters and cups-filters?
On 08/06/2014 19:15, Andreas K. Huettel wrote: Am Sonntag, 8. Juni 2014, 17:48:09 schrieb Alan Mackenzie: . What is all this trying to tell me? I've tried for over an hour to make sense of it, but my eyes just glaze over. My best guess is that cups-filters and foomatic-filters are packages which can't be installed together. But I _need_ foomatic-filters - otherwise my printer doesn't print. Or do I? cups-filters seems to be needed by cups. What _are_ cups-filters and foomatic-filters? emerge -s is little help here. Why do I need both of them? * cups-filters is a former part of cups that provides file format conversions (among other things). Basically it (also) makes sure that everything is internally converted to PDF. It's not part of CUPS (as maintained by Apple) anymore, but hard-required by CUPS on Linux (and maintained by the Linux Foundation). * foomatic-filters is a set of printer drivers, basically. * Some time ago the cups-filters maintainers took over maintainership of the foomatic-filters part for CUPS as well, and integrated it cleanly into cups- filters. That's the reason for the blocker; recent cups-filters contain the newest foomatic code available. The former separate foomatic-filters package is now unmaintained. So, we have the following possibilities for installation: 1) normal CUPS user, recommended, this is what comes by default (unless you do something stupid such as USE=-*) net-print/cups net-print/cups-filters[foomatic] 2) NOT recommended, dead code, unmaintained: net-print/cups net-print/cups-filters[-foomatic] net-print/foomatic-filters 3) for the stone age people out there, NOT recommended, dead code, unmaintained: any other printing system, e.g. lprng net-print/foomatic-filters So, what's wrong in your case? No idea, but after longish not-updating things do get hard for emerge to unravel. My recommendation is, since foomatic- filters and cups-filters are only needed for printing and emerge runs fine without them, force-remove both and let emerge figure out the right package set from scratch. [This basically works with any blocker as a last resort, but can be *very* dangerous for packages that are needed by the core system. You definitely don't want to remove gcc or glibc this way, for example. :)] emerge -aC net-print/cups-filters net-print/foomatic-filters emerge -uDNavt --backtrack=100 world Cheers, Andreas Good post! For Alan Mackenzie's benefit, a little back story: The whole topic of printing is a mess, no single mere mortal can wrap their wits around it. Long long ago a printer was a piece of hardware you plugged into a serial or parallel port, the kernel found it and you were good to go. Whoopee! Because more than one user could use the printer and this causes conflicts, print servers were written: the server controlled the printer hardware and you submitted your print job to the server, and that took care of all the messy parts. To do it over the network was just as easy, modify the print server to also listen on a network port. This server was the classic lp suite of tools. Many years ago, HP developed a fancy printing language for their laser printers called PostScript[1]. Think of it as a giant image format, it doesn't describe what the printed page looks like, it really is simple code that tells the printer how to print the page, including graphics and such. And so the era of complicated drivers was begun. These laser printers needed gobs of memory and big cpus to deal with PostScript, in the 386 era it was common to have a printer much more powerful than your computer. Enter other vendors and Windows. Just like with sound cards, vendors wrote their own drivers adding features done in software. This makes sense is you can't get PostScript to do double-sided printing or scale down so two pages fit on one page, doesn't make so much sense if you just want to avoid paying HP a PostScript license. After a while, HP got around to updating PostScript (or maybe it was Apple's code all long - I forget...) and called it PCL (Printer Control Language), needing new drivers. meanwhile, printers shifted over to USB away from parallel ports and this needed new drivers. Plus there's two way to do it: do the USB part of the printing in userspace and only use the kernel for regular USB work, or put the whole thing in the kernel. Needing more drivers. last I looked, there were still some serious issues with the options to have it all in the kernel. On the print server side, the devs were getting real busy. We had classic lp, then came lprng, then something else I forget and finally an upstart crowd wrote CUPS (Common Unix Printing System), eventually bought by Apple. Ironically, there's now nothing common about it and it's for iOS not Unix. Such is life. With the latest major version update Apple ripped out all the bits we find so useful and still declare the software is for
Re: [gentoo-user] What's with foomatic-filters and cups-filters?
Alan McKinnon wrote: And I haven't even touched on CUPS' feature that requires you to delete and re-add back all your printers after any remerge. Ask Dale about this, he's the resident expert and he's even figured out how to get hplip to work. Every time I upgrade CUPS or hplip, I go to a Konsole and type in hp-setup as root. A window pops up and I just set the printer up again, it's GUI based. So far, that has worked. Don't jinx it tho. lol If needed, I go to my web browser to CUPS and delete the printer first. Since I ran out of ink, I haven't printed in a while however, the same works on my brothers puter and he runs Kubuntu. Well, was my brothers anyway. :/ Dale :-) :-) -- I am only responsible for what I said ... Not for what you understood or how you interpreted my words!
Re: [gentoo-user] What's with foomatic-filters and cups-filters?
On 06/08/2014 03:08 PM, Dale wrote: Every time I upgrade CUPS or hplip, I go to a Konsole and type in hp-setup as root. A window pops up and I just set the printer up again, it's GUI based. So far, that has worked. Don't jinx it tho. lol Yep, same here. I read a lot of horror stories getting it to work, and when I first installed hplip it didn't do anything until I googled and found I had to run hp-setup with elevated privileges. I haven't had any issues printing - hplip prints great with my CP1025nw. If needed, I go to my web browser to CUPS and delete the printer first. I hate updating cups and hplip, I've masked newer versions and will only update when I really have to. (Like another package needing a new version of something-or-other.) Since I ran out of ink, I haven't printed in a while however, the same works on my brothers puter and he runs Kubuntu. Well, was my brothers anyway. :/ I'm using hplip on Ubuntu, Mint, and Gentoo. Dan
Re: [gentoo-user] What's with foomatic-filters and cups-filters?
On 09/06/2014 00:08, Dale wrote: Alan McKinnon wrote: And I haven't even touched on CUPS' feature that requires you to delete and re-add back all your printers after any remerge. Ask Dale about this, he's the resident expert and he's even figured out how to get hplip to work. Every time I upgrade CUPS or hplip, I go to a Konsole and type in hp-setup as root. A window pops up and I just set the printer up again, it's GUI based. So far, that has worked. Don't jinx it tho. lol printing? printing? who mentioned printing? sure wasn't me, I remember all too well what happened that time we mentioned HAL and you were around that was not good and none of us want *that* again so we are very careful in what we say to keep the jinx monster away :-) You've been quite lately though, what happened? Surely it can't be that Gentoo finally stopped messing with you and started behaving itself? hahahaha Alan made a funny -- Alan McKinnon alan.mckin...@gmail.com
[gentoo-user] numlock script and systemd
Hi. I would like to use my /etc/init.d/numlock script with systemd, but
systemctl start just says warning numlock is already starting. I
thought that if a service was not there it would use something in
/etc/init.d and create one, so what am I doing wrong?
Here is the script
--cut here --
#!/sbin/runscript
# Copyright (c) 2007-2009 Roy Marples r...@marples.name
# Released under the 2-clause BSD license.
description=Turns numlock on for the consoles.
ttyn=${rc_tty_number:-${RC_TTY_NUMBER:-12}}
depend()
{
need localmount
keyword -openvz -prefix -vserver -lxc
}
_setleds()
{
[ -z $1 ] return 1
local dev=/dev/tty t= i=1 retval=0
[ -d /dev/vc ] dev=/dev/vc/
while [ $i -le $ttyn ]; do
setleds -D $1num $dev$i || retval=1
: $(( i += 1 ))
done
return $retval
}
start()
{
ebegin Enabling numlock on ttys
_setleds +
eend $? Failed to enable numlock
}
stop()
{
ebegin Disabling numlock on ttys
_setleds -
eend $? Failed to disable numlock
}
- cut here --
It does work under openrc.
Thanks in advance for any suggestions.
--
Your life is like a penny. You're going to lose it. The question is:
How do
you spend it?
John Covici
cov...@ccs.covici.com
Re: [gentoo-user] numlock script and systemd
On Sun, Jun 8, 2014 at 9:39 PM, cov...@ccs.covici.com wrote: Hi. I would like to use my /etc/init.d/numlock script with systemd, but systemctl start just says warning numlock is already starting. I thought that if a service was not there it would use something in /etc/init.d and create one, so what am I doing wrong? That only works for normal SysV scripts. OpenRC scripts are similar, but different enough so that systemd on Gentoo doesn't even tries to load them. To do what you want, there are several ways: for example you can create a service for each terminal and enable all of them, or you can create a template service, and create instances for all your terminals. However, in systemd all the VTs are not started until you switch to them (except the first one), sou you have the start the services for each terminal *after* getty@ttyN.service, with N=1,2,... I think the easiest option is for you to override getty@.service, by creating /etc/systemd/systemd/getty@.service.d/numlock.conf, and putting: ExecStartPost=setleds -D +num /dev/%I in int. Then you only do systemctl --system daemon-reload, and if everything works, you should have numlock activated in each VT when you switch to them. Check man 5 systemd.unit; the proper explanation for all the options is given there. Regards. -- Canek Peláez Valdés Profesor de asignatura, Facultad de Ciencias Universidad Nacional Autónoma de México
Re: [gentoo-user] numlock script and systemd
Canek Peláez Valdés can...@gmail.com wrote: On Sun, Jun 8, 2014 at 9:39 PM, cov...@ccs.covici.com wrote: Hi. I would like to use my /etc/init.d/numlock script with systemd, but systemctl start just says warning numlock is already starting. I thought that if a service was not there it would use something in /etc/init.d and create one, so what am I doing wrong? That only works for normal SysV scripts. OpenRC scripts are similar, but different enough so that systemd on Gentoo doesn't even tries to load them. To do what you want, there are several ways: for example you can create a service for each terminal and enable all of them, or you can create a template service, and create instances for all your terminals. However, in systemd all the VTs are not started until you switch to them (except the first one), sou you have the start the services for each terminal *after* getty@ttyN.service, with N=1,2,... I think the easiest option is for you to override getty@.service, by creating /etc/systemd/systemd/getty@.service.d/numlock.conf, and putting: ExecStartPost=setleds -D +num /dev/%I in int. Then you only do systemctl --system daemon-reload, and if everything works, you should have numlock activated in each VT when you switch to them. Check man 5 systemd.unit; the proper explanation for all the options is given there. Thanks, I have put it in, we shall see what happens on the next reboot. -- Your life is like a penny. You're going to lose it. The question is: How do you spend it? John Covici cov...@ccs.covici.com
