[gentoo-user] grub2-mkconfig vs. xen
Hi, I have a box running a xen-domu. If I want to use grub2-mkconfig, I learned that I have to put the kernel-.config into /etc/kernels. However this does not create an entry which includes the initrd (it creates one without xen though). How do I achieve this? Also I only want entries with xen and I need to pass some params to xen as well as the linux kernel. Which entries in /etc/default/grub do that? Regards, Konstantin -- Dipl-Inf. Konstantin Agouros aka Elwood Blues. Internet: elw...@agouros.de Altersheimerstr. 1, 81545 Muenchen, Germany. Tel +49 89 69370185 "There is no 'dead' in team!" - Sameen Shaw
[gentoo-user] world file - where has it gone?
Hi, I discovered that the file /var/lib/portage/world contains only a few lines where it contained hundreds of files before. Has the information previously kept in this file been moved to some other place (database?)? Many thanks for a hint, Helmut
[gentoo-user] Re: world file - where has it gone?
Helmut Jarausch: >I discovered that the file /var/lib/portage/world >contains only a few lines where it contained hundreds of files >before. On my mostly stable Gentoo using portage 2.1.14 I have at the moment hafi@i5_64 ~ $ wc -l /var/lib/portage/world 167 /var/lib/portage/world Hartmut
Re: [gentoo-user] grub2-mkconfig vs. xen
On Sun, 4 Jan 2015 10:44:52 + (UTC), Konstantinos Agouros wrote: > How do I achieve this? Also I only want entries with xen and I need > to pass some params to xen as well as the linux kernel. Which entries > in /etc/default/grub do that? chmod -x /etc/grub.d/10_linux to prevent the creation of standard linux stanzas. The configuration section of info grub details the settings in defaults, you want to look at GRUB_CMDLINE_XEN and GRUB_CMDLINE_XEN_DEFAULT. -- Neil Bothwick The Japanese call us lazy, but at least we cook our fish! pgppC3d9nKj9X.pgp Description: OpenPGP digital signature
[gentoo-user] Re: world file - where has it gone?
Hartmut Figge: >On my mostly stable Gentoo using portage 2.1.14 I have at the moment 2.2.14 Hartmut
Re: [gentoo-user] NSA SELinux kernel support
> I was wondering if there was any harm in disabling the NSA SELinux support > in my gentoo-sources based kernel. There is no harm, but if you were interested a lot of packages come with policies by default. Currently there is no support for SELinux in Gentoo for the vast majority of desktop applications. It is a little bit of work to get anything nonfunctional working. There are additional modes where you can simply run your user as unconfined and any services will be restricted by SELinux. grsecurity's RBAC is an alternative where you simply let it generate a policy based on what it sees you use. Notably, Fedora and CentOS enable SELinux by default. > SELinux is the only one I've had a bit of experience with - I run CentOS > (SELinux is enabled by default) for some personal-use-only services that > I want to run without dealing with Gentoo. My first step in a CentOS > install is to disable SELinux (and the firewall, hehe) to avoid dealing > with the pain of wading through documentation for hours on end. http://stopdisablingselinux.com/ - your distribution probably comes with policies for everything you want to install, anyway...
Re: [gentoo-user] Re: Libreoffice PDF conversion problem : solved
150102 Philip Webb wrote: > 150101 walt wrote: >> On 12/31/2014 04:40 AM, Philip Webb wrote: >>> I want to revise the 1st page of a PDF on my Internet site. >>> The revision I created today has mangled some of the punctuation. >>> Apostrophes & extended dashes are mangled, >>> even when I enter a completely new .odt file. >>> My egs are at http://www.chass.utoronto.ca/~purslow/test/ . >>> The problem occurs with Okular, Mupdf & Firefox's PDF viewer. >>> I've submitted LibreOffice Bug 87903 . Can anyone else reproduce this ? >> I'm running 4.3.5.2, the testing version on ~amd64, >> and I don't see the same export error, >> when I use the two test .odt files downloaded from your bug report. > Ah, interesting ! -- I'm using 4.3.4.1 , so perhaps it's been corrected. I've moved up to 4.3.5.2 & the problem doesn't occur : thanks again. -- ,, SUPPORT ___//___, Philip Webb ELECTRIC /] [] [] [] [] []| Cities Centre, University of Toronto TRANSIT`-O--O---' purslowatchassdotutorontodotca
[gentoo-user] Re: pdf viewer
On 2015-01-03, the wrote: > > On 03/01/15 08:15, lee wrote: >> Hi, >> >> what do you as PDF viewer? Most of the time, I use zathura with libmupdf; sometimes I will use evince when I need to view documents with libpoppler, but evince is not my main document viewer ever since GNOME started playing games with UI design. Both zathura and evince allow me to see a wider range of documents (including DeJaVU and Adobe PostScript), and also have a "reverse video" feature. > I was using evince, now moved to zathura. Both are as slow as hell > though, so I'm open for faster alternatives. Are you using zathura with libmupdf or with libpoppler? You can always try one, then the other, and see which one do you prefer. -- Nuno Silva (aka njsg) Helsinki, Finland
[gentoo-user] Slow startx - Why does hostname -f hang?
Hi list While hunting down slow startx times (12-16s instead the usual 2-5s) I found that "hostname -f" tries to resolve the hostname. This is a slow process (timeout?): urs@cadd ~ $ time hostname -v gethostname()=`cadd' cadd real0m0.001s user0m0.000s sys 0m0.001s urs@cadd ~ $ time hostname -v -f gethostname()=`cadd' Resolving `cadd' ... Result: h_name=`cadd' Result: h_addr_list=`127.0.0.1' cadd real0m10.011s user0m0.000s sys 0m0.001s urs@cadd ~ $ man hostname ...snip... -f, --fqdn, --long Display the FQDN (Fully Qualified Domain Name). A FQDN consists of a short host name and the DNS domain name. Unless you are using bind or NIS for host lookups you can change the FQDN and the DNS domain name (which is part of the FQDN) in the /etc/hosts file. ...snip... My /etc/hosts.conf orders hosts before bind: urs@cadd ~ $ grep order /etc/host.conf # recognized are order, trim, mdns, multi, nospoof, spoof, and reorder. order hosts, bind I do not run any NIS or bind services. I solved the slow startx by replacing "hostname -f" by "hostname" in the /usr/bin/startx script, but still feel that I'm missing the real cause for the slow hostname lookup. I do not have a DNS domain name definition in /etc/hosts. I do run networkmanager. Any hints where to search? What is the recommended way for /etc/hosts? I'm at a simple home network, behind a NAT cable modem, and do not have a dns domain name. Thanks for any hints. Urs
Re: [gentoo-user] NSA SELinux kernel support
On 01/04/2015 09:47 AM, Sid S wrote: > >> SELinux is the only one I've had a bit of experience with - I run CentOS >> (SELinux is enabled by default) for some personal-use-only services that >> I want to run without dealing with Gentoo. My first step in a CentOS >> install is to disable SELinux (and the firewall, hehe) to avoid dealing >> with the pain of wading through documentation for hours on end. > http://stopdisablingselinux.com/ - your distribution probably comes > with policies for everything you want to install, anyway... > > > Thanks for this link - I'll watch that video later this afternoon I think. Alec
Re: [gentoo-user] Slow startx - Why does hostname -f hang?
On Sunday 04 Jan 2015 16:06:50 Urs Schütz wrote: > Hi list > > While hunting down slow startx times (12-16s instead the usual 2-5s) I > found that "hostname -f" tries to resolve the hostname. This is a slow > process (timeout?): > > urs@cadd ~ $ time hostname -v > gethostname()=`cadd' > cadd > > real0m0.001s > user0m0.000s > sys 0m0.001s > > urs@cadd ~ $ time hostname -v -f > gethostname()=`cadd' > Resolving `cadd' ... > Result: h_name=`cadd' > Result: h_addr_list=`127.0.0.1' > cadd > > real0m10.011s > user0m0.000s > sys 0m0.001s > > urs@cadd ~ $ man hostname > ...snip... > -f, --fqdn, --long >Display the FQDN (Fully Qualified Domain Name). A FQDN > consists >of a short host name and the DNS domain name. Unless > you are >using bind or NIS for host lookups you can change the > FQDN and >the DNS domain name (which is part of the FQDN) > in the >/etc/hosts file. > ...snip... > > My /etc/hosts.conf orders hosts before bind: > urs@cadd ~ $ grep order /etc/host.conf > # recognized are order, trim, mdns, multi, nospoof, spoof, and reorder. > order hosts, bind > > I do not run any NIS or bind services. > > I solved the slow startx by replacing "hostname -f" by "hostname" in the > /usr/bin/startx script, but still feel that I'm missing the real cause > for the slow hostname lookup. > > I do not have a DNS domain name definition in /etc/hosts. > I do run networkmanager. > Any hints where to search? > > What is the recommended way for /etc/hosts? I'm at a simple home > network, behind a NAT cable modem, and do not have a dns domain name. > > Thanks for any hints. > Urs In my /etc/hosts I have something like this: # IPv4 and IPv6 localhost aliases 127.0.0.1 cad.homeLAN localhost cad ::1 cad.homeLAN localhost cad See if this solves your problem. PS. I remember a thread in this M/L about the correct way to configure a domain name and why how the Gentoo Handbook was wrong. I think there was a bug opened about it, too. -- Regards, Mick signature.asc Description: This is a digitally signed message part.
Re: [gentoo-user] Slow startx - Why does hostname -f hang?
Am Sonntag, 04.01.2015 um 14:06 schrieb Urs Schütz : > Hi list > > While hunting down slow startx times (12-16s instead the usual 2-5s) > I found that "hostname -f" tries to resolve the hostname. This is a > slow process (timeout?): Normally this is a fast process. See my results below. > urs@cadd ~ $ time hostname -v > gethostname()=`cadd' > cadd > > real0m0.001s > user0m0.000s > sys 0m0.001s user@puter ~ $ time hostname -v gethostname()=`puter' puter real0m0.003s user0m0.000s sys 0m0.000s > urs@cadd ~ $ time hostname -v -f > gethostname()=`cadd' > Resolving `cadd' ... > Result: h_name=`cadd' > Result: h_addr_list=`127.0.0.1' ^ Does your computer only have an IP address for your localhost? > cadd > > real0m10.011s > user0m0.000s > sys 0m0.001s user@puter ~ $ time hostname -v -f gethostname()=`puter' Resolving `puter' ... Result: h_name=`puter.local' Result: h_aliases=`puter' Result: h_addr_list=`192.168.44.32' puter.local real0m0.005s user0m0.003s sys 0m0.003s > What is the recommended way for /etc/hosts? I'm at a simple home > network, behind a NAT cable modem, and do not have a dns domain name. What is the content of your /etc/hosts file? This is my hostfile: 127.0.0.1 localhost ::1 localhost 192.168.44.32 puter.local puter Regards wabe
Re: [gentoo-user] Slow startx - Why does hostname -f hang?
On 01/04/15 15:20, Mick wrote: On Sunday 04 Jan 2015 16:06:50 Urs Schütz wrote: Hi list While hunting down slow startx times (12-16s instead the usual 2-5s) I found that "hostname -f" tries to resolve the hostname. This is a slow process (timeout?): urs@cadd ~ $ time hostname -v gethostname()=`cadd' cadd real0m0.001s user0m0.000s sys 0m0.001s urs@cadd ~ $ time hostname -v -f gethostname()=`cadd' Resolving `cadd' ... Result: h_name=`cadd' Result: h_addr_list=`127.0.0.1' cadd real0m10.011s user0m0.000s sys 0m0.001s urs@cadd ~ $ man hostname ...snip... -f, --fqdn, --long Display the FQDN (Fully Qualified Domain Name). A FQDN consists of a short host name and the DNS domain name. Unless you are using bind or NIS for host lookups you can change the FQDN and the DNS domain name (which is part of the FQDN) in the /etc/hosts file. ...snip... My /etc/hosts.conf orders hosts before bind: urs@cadd ~ $ grep order /etc/host.conf # recognized are order, trim, mdns, multi, nospoof, spoof, and reorder. order hosts, bind I do not run any NIS or bind services. I solved the slow startx by replacing "hostname -f" by "hostname" in the /usr/bin/startx script, but still feel that I'm missing the real cause for the slow hostname lookup. I do not have a DNS domain name definition in /etc/hosts. I do run networkmanager. Any hints where to search? What is the recommended way for /etc/hosts? I'm at a simple home network, behind a NAT cable modem, and do not have a dns domain name. Thanks for any hints. Urs In my /etc/hosts I have something like this: # IPv4 and IPv6 localhost aliases 127.0.0.1 cad.homeLAN localhost cad ::1 cad.homeLAN localhost cad See if this solves your problem. PS. I remember a thread in this M/L about the correct way to configure a domain name and why how the Gentoo Handbook was wrong. I think there was a bug opened about it, too. Setting up /etc/hosts exactly as above did resolve the slow hostname -f lookup. urs@cadd ~ $ time hostname -v -f gethostname()=`cadd' Resolving `cadd' ... Result: h_name=`cadd.homeLAN' Result: h_aliases=`localhost' Result: h_aliases=`cadd' Result: h_addr_list=`::1' cadd.homeLAN real0m0.001s user0m0.000s sys 0m0.000s :-) Thanks for your help. Urs
Re: [gentoo-user] SOLVED: Slow startx - Why does hostname -f hang?
On 01/04/15 15:27, waben...@gmail.com wrote: Am Sonntag, 04.01.2015 um 14:06 schrieb Urs Schütz : Hi list While hunting down slow startx times (12-16s instead the usual 2-5s) I found that "hostname -f" tries to resolve the hostname. This is a slow process (timeout?): Normally this is a fast process. See my results below. urs@cadd ~ $ time hostname -v gethostname()=`cadd' cadd real0m0.001s user0m0.000s sys 0m0.001s user@puter ~ $ time hostname -v gethostname()=`puter' puter real0m0.003s user0m0.000s sys 0m0.000s urs@cadd ~ $ time hostname -v -f gethostname()=`cadd' Resolving `cadd' ... Result: h_name=`cadd' Result: h_addr_list=`127.0.0.1' ^ Does your computer only have an IP address for your localhost? cadd real0m10.011s user0m0.000s sys 0m0.001s user@puter ~ $ time hostname -v -f gethostname()=`puter' Resolving `puter' ... Result: h_name=`puter.local' Result: h_aliases=`puter' Result: h_addr_list=`192.168.44.32' puter.local real0m0.005s user0m0.003s sys 0m0.003s What is the recommended way for /etc/hosts? I'm at a simple home network, behind a NAT cable modem, and do not have a dns domain name. What is the content of your /etc/hosts file? This is my hostfile: 127.0.0.1 localhost ::1 localhost 192.168.44.32 puter.local puter Regards wabe Hi Wabe IP changes dynamically, is assigned by Networkmanager with dhclient use flag. I changed /etc/hosts as Mick/Michael pointed out in an other reply, and this solved the slow response. Here the relevant part of the corrected, working /etc/hosts: # IPv4 and IPv6 localhost aliases 127.0.0.1 cadd.homeLAN localhost ::1 cadd.homeLAN localhost cadd Thanks for your time looking into it and the hints. Urs
Re: [gentoo-user] NSA SELinux kernel support
Sid S writes: > your distribution probably comes > with policies for everything you want to install, anyway... ...until it doesn't, and then what? I attempted a full conversion a few months back, and was ready to make some commitment to getting SELinux to work on my personal laptop. I got as far as Permissive mode, with a firehose of access violations in the auditd log. I had written a couple of scrappy policies to authorize a few small one-off violations, with the help of audit2allow, but the firehose was still gushing. I use offlineimap for fetching mail, which doesn't have a policy. Now, if I ever wanted to switch from Permissive to Enforcing, I was required, as an absolute SELinux n00b, to write a full policy for a non-trivial mail application. This is when I turned around. I could have half-assed it with audit2allow, but security-wise that's a cop-out. Inevitably, there will always be some program I want to use with no existing policy, and I'll constantly have this problem. I realized that my personal workstation is a place I like to try lots of software (don't we all like that about Linux?), and SELinux can be a big wet blanket on the fun at any time. I'd like to find a middle ground, and it might be Targeted mode (I was attempting Strict). Or, it might be a different system like AppArmor. -- Erik Mackdanz
Re: [gentoo-user] Slow startx - Why does hostname -f hang?
On Sunday 04 Jan 2015 18:45:38 Urs Schütz wrote: > On 01/04/15 15:20, Mick wrote: > > In my /etc/hosts I have something like this: > > > > # IPv4 and IPv6 localhost aliases > > 127.0.0.1 cad.homeLAN localhost cad > > > > ::1 cad.homeLAN localhost cad > > > > See if this solves your problem. > > > > PS. I remember a thread in this M/L about the correct way to configure a > > domain name and why how the Gentoo Handbook was wrong. I think there was > > a bug opened about it, too. > > Setting up /etc/hosts exactly as above did resolve the slow hostname -f > lookup. > > urs@cadd ~ $ time hostname -v -f > gethostname()=`cadd' > Resolving `cadd' ... > Result: h_name=`cadd.homeLAN' > Result: h_aliases=`localhost' > Result: h_aliases=`cadd' > Result: h_addr_list=`::1' > cadd.homeLAN > > real0m0.001s > user0m0.000s > sys 0m0.000s > > :-) Thanks for your help. You're welcome! :-) Please note that I assumed the name of your host PC is "cad" and the arbitrary name of your local domain is "homeLAN". You may want to change these to suit your needs. -- Regards, Mick signature.asc Description: This is a digitally signed message part.
Re: [gentoo-user] world file - where has it gone?
On Sun, Jan 04, 2015 at 01:19:03PM +0100, Helmut Jarausch wrote: > Hi, > I discovered that the file /var/lib/portage/world > contains only a few lines where it contained hundreds of files > before. As far as I’m aware, if you say `emerge foo`, then "foo" is added to that file. But a long time ago I switched to manually managing sets (essentials, kde, office, etc.). If you add -1 to the emerge command, then foo is not added to world. Perhaps you started using that argument. -- Gruß | Greetings | Qapla’ Please do not share anything from, with or about me with any social network. “If a ship with 1000 investment bankers sinks in heavy seas, it is a tragedy. If only one of them can swim, it is a disaster.” – Urban Priol signature.asc Description: Digital signature
Re: [gentoo-user] grub2-mkconfig vs. xen
On 2015-01-04 11:44, Konstantinos Agouros wrote: Hi, I have a box running a xen-domu. If I want to use grub2-mkconfig, I learned that I have to put the kernel-.config into /etc/kernels. However this does not create an entry which includes the initrd (it creates one without xen though). How do I achieve this? Also I only want entries with xen and I need to pass some params to xen as well as the linux kernel. Which entries in /etc/default/grub do that? Regards, Konstantin You may put your kernel configuration in /boot as well. The important thing is to have xen enabled in that configuration - that is what grub looks for. It only creates xen entries if xen is enabled. Also the name must match - for example config-*, kernel-* and so on. You can can set to boot xen by default with grub2-set-default (specify the menuentry) - it's written to grubenv. Before that, set GRUB_DEFAULT="saved" to /etc/default/grub.
Re: [gentoo-user] world file - where has it gone?
On 2015-01-04 13:19, Helmut Jarausch wrote: Hi, I discovered that the file /var/lib/portage/world contains only a few lines where it contained hundreds of files before. Has the information previously kept in this file been moved to some other place (database?)? Many thanks for a hint, Helmut Do you specify the "oneshot" parameter when you emerge packages? Does it contain the packages you really need (not the dependencies)?
[gentoo-user] lxc.network.flags
Hi, what does 'lxc.network.flags=up' actually do/mean? 1.) makes the interface available in the container 2.) does 1.) and also configures the containers' interface 3.) something else (what?) In case of 2.), how do I set the gateway address, nameserver and domain in the containers' config? I suppose I need to omit the 'lxc.network.flags' and to make the container set up the interface itself. Will the interface be available in the container when it's not brought up through the containers' config? -- Again we must be afraid of speaking of daemons for fear that daemons might swallow us. Finally, this fear has become reasonable.
[gentoo-user] setxkbmap -option ctrl:nocaps
Hi, I'm trying to use the CapsLock key as control key. With 'setxkbmap -option ctrl:nocaps' that works --- until I use 'xmodmap' to load my keymap. Once the keymap is loaded, the CapsLock key again works as CapsLock and not as control. This worked fine with other distributions. Is there anything special to Gentoo which breaks this? Section "InputClass" Identifier "keyboard-all" Driver "evdev" Option "XkbLayout" "de" Option "XkbModel" "pc105" Option "XkbRules" "xorg" Option "XkbOptions" "ctrl:nocaps,terminate:ctrl_alt_bksp" MatchIsKeyboard "on" EndSection ~/.xinitrc: [...] setxkbmap -option ctrl:nocaps xmodmap ~/.Xmodmap [...] ~/.Xmodmap: [...] keycode 37 = Control_L NoSymbol Control_L [...] -- Again we must be afraid of speaking of daemons for fear that daemons might swallow us. Finally, this fear has become reasonable.
[gentoo-user] PYTHON_CFLAGS is invalid for python-r1 suite
When installing app-office/libreoffice-4.2.8.2, I got: >>> Messages generated by process 3014 on 2015-01-04 22:21:49 EET for package app-office/libreoffice-4.2.8.2: ERROR: setup ERROR: app-office/libreoffice-4.2.8.2::gentoo failed (setup phase): PYTHON_CFLAGS is invalid for python-r1 suite, please take a look @ https://wiki.gentoo.org/wiki/Project:Python/Python.eclass_conversion#PYTHON_CFLAGS I did have a look at the URL specified, but couldn't make head nor tail of it. Is there a fix for that? Or do I just report this to the package maintainer? Thanks.
Re: [gentoo-user] PYTHON_CFLAGS is invalid for python-r1 suite
Am Sonntag, 4. Januar 2015, 21:46:41 schrieb Alexander Kapshuk: > When installing app-office/libreoffice-4.2.8.2, I got: > >>> Messages generated by process 3014 on 2015-01-04 22:21:49 EET for > > package app-office/libreoffice-4.2.8.2: > > ERROR: setup > ERROR: app-office/libreoffice-4.2.8.2::gentoo failed (setup phase): > PYTHON_CFLAGS is invalid for python-r1 suite, please take a look @ > https://wiki.gentoo.org/wiki/Project:Python/Python.eclass_conversion#PYTHON > _CFLAGS > > I did have a look at the URL specified, but couldn't make head nor tail of > it. > > Is there a fix for that? Or do I just report this to the package > maintainer? > > Thanks. What exact command are you using to install libreoffice? Is PYTHON_CFLAGS set in your make.conf? -- Andreas K. Huettel Gentoo Linux developer (council, kde) dilfri...@gentoo.org http://www.akhuettel.de/ signature.asc Description: This is a digitally signed message part.
Re: [gentoo-user] PYTHON_CFLAGS is invalid for python-r1 suite
Am Sonntag, 4. Januar 2015, 22:30:54 schrieb Andreas K. Huettel: > Am Sonntag, 4. Januar 2015, 21:46:41 schrieb Alexander Kapshuk: > > When installing app-office/libreoffice-4.2.8.2, I got: > > >>> Messages generated by process 3014 on 2015-01-04 22:21:49 EET for > > > > package app-office/libreoffice-4.2.8.2: > > > > ERROR: setup > > > > ERROR: app-office/libreoffice-4.2.8.2::gentoo failed (setup phase): > > PYTHON_CFLAGS is invalid for python-r1 suite, please take a look @ > > > > https://wiki.gentoo.org/wiki/Project:Python/Python.eclass_conversion#PYTH > > ON _CFLAGS > > > > I did have a look at the URL specified, but couldn't make head nor tail > > of it. > > > > Is there a fix for that? Or do I just report this to the package > > maintainer? > > > > Thanks. > > What exact command are you using to install libreoffice? > Is PYTHON_CFLAGS set in your make.conf? Never mind, we found the problem. Libreoffice can be emerged and installed fine from sources, but every binpackage that you yourself generate from sources right now cannot be installed again... :/ (Does NOT affect libreoffice-bin.) Working on it. -- Andreas K. Huettel Gentoo Linux developer (council, kde) dilfri...@gentoo.org http://www.akhuettel.de/ signature.asc Description: This is a digitally signed message part.
[gentoo-user] printer problem solved by new flag
I've just successfully got my HP DJ-2510 printer to work after a struggle. I updated Hplip recently & there's a new flag : USE="hpcups" ; in the printer dialog at Port 631, this needs to be chosen : when I used the 'hpijs' filter, it refused to print with "filter failed", but when I chose 'hpcups' it went thro' properly. I'm reporting this to help others who may run into the roadblock. -- ,, SUPPORT ___//___, Philip Webb ELECTRIC /] [] [] [] [] []| Cities Centre, University of Toronto TRANSIT`-O--O---' purslowatchassdotutorontodotca
Re: [gentoo-user] PYTHON_CFLAGS is invalid for python-r1 suite
> > Am Sonntag, 4. Januar 2015, 21:46:41 schrieb Alexander Kapshuk: > > > When installing app-office/libreoffice-4.2.8.2, I got: > > > >>> Messages generated by process 3014 on 2015-01-04 22:21:49 EET for > > > > > > package app-office/libreoffice-4.2.8.2: > > > > > > ERROR: setup > > > > > > ERROR: app-office/libreoffice-4.2.8.2::gentoo failed (setup phase): > > > PYTHON_CFLAGS is invalid for python-r1 suite, please take a look @ > > > > > > https://wiki.gentoo.org/wiki/Project:Python/Python.eclass_conversion#PY > > > TH ON _CFLAGS > > > > > Never mind, we found the problem. > Working on it. http://dilfridge.blogspot.de/2015/01/broken-app-officelibreoffice-binary.html -- Andreas K. Huettel Gentoo Linux developer (council, kde) dilfri...@gentoo.org http://www.akhuettel.de/ signature.asc Description: This is a digitally signed message part.
Re: [gentoo-user] setxkbmap -option ctrl:nocaps
On Sunday 04 Jan 2015 19:53:37 lee wrote: > Hi, > > I'm trying to use the CapsLock key as control key. With 'setxkbmap > -option ctrl:nocaps' that works --- until I use 'xmodmap' to load my > keymap. Once the keymap is loaded, the CapsLock key again works as > CapsLock and not as control. > > This worked fine with other distributions. Is there anything special to > Gentoo which breaks this? > > > Section "InputClass" > Identifier "keyboard-all" > Driver "evdev" > Option "XkbLayout" "de" > Option "XkbModel" "pc105" > Option "XkbRules" "xorg" > Option "XkbOptions" "ctrl:nocaps,terminate:ctrl_alt_bksp" > MatchIsKeyboard "on" > EndSection What happens if you remove "ctrl:nocaps" above? > ~/.xinitrc: > [...] > setxkbmap -option ctrl:nocaps > xmodmap ~/.Xmodmap > [...] > > > ~/.Xmodmap: > [...] > keycode 37 = Control_L NoSymbol Control_L > [...] and why do you need to specify this in .Xmodap? -- Regards, Mick signature.asc Description: This is a digitally signed message part.
Re: [gentoo-user] PYTHON_CFLAGS is invalid for python-r1 suite
On Mon, Jan 5, 2015 at 2:01 AM, Andreas K. Huettel wrote: > > > Am Sonntag, 4. Januar 2015, 21:46:41 schrieb Alexander Kapshuk: > > > > When installing app-office/libreoffice-4.2.8.2, I got: > > > > >>> Messages generated by process 3014 on 2015-01-04 22:21:49 EET for > > > > > > > > package app-office/libreoffice-4.2.8.2: > > > > > > > > ERROR: setup > > > > > > > > ERROR: app-office/libreoffice-4.2.8.2::gentoo failed (setup phase): > > > > PYTHON_CFLAGS is invalid for python-r1 suite, please take a look @ > > > > > > > > > https://wiki.gentoo.org/wiki/Project:Python/Python.eclass_conversion#PY > > > > TH ON _CFLAGS > > > > > > > > Never mind, we found the problem. > > Working on it. > > > http://dilfridge.blogspot.de/2015/01/broken-app-officelibreoffice-binary.html > > -- > Andreas K. Huettel > Gentoo Linux developer (council, kde) > dilfri...@gentoo.org > http://www.akhuettel.de/ > Understood. Thanks. Sorry I didn't specify in the first place that I it was a binary package I had generated from the portage tree that I was attempting to install. I have synced my portage tree to pull in the updated /usr/portage/eclass/python-single-r1.eclass. I will try unsetting PYTHON_CFLAGS in the environment file contained in the binary package generated as suggested here, https://bugs.gentoo.org/show_bug.cgi?id=534726, and see if that works. Thanks.
