Re: [arch-general] Android support in Linux Arch
Hi, Am 17.04.2014 04:22, schrieb GSC: You can just chown /opt/android-sdk and it will be easier to install api. Yeah, although I don't like the idea to mess around in /opt. Maybe it would be possible to introduce an android group, so it would be sufficient to add my user to this group? especially some huge package like android-ndk. I don' t know why you only consider non-binary packages. +1. android-ndk is a pain in the ass to compile, especially on older hardware, where it takes not only long but might run into memory restrictions, because it takes up so much RAM. Best regards, Karol Babioch signature.asc Description: OpenPGP digital signature
Re: [arch-general] Android support in Linux Arch
On 17/04/14 02:20 AM, Karol Babioch wrote: Hi, Am 17.04.2014 04:22, schrieb GSC: You can just chown /opt/android-sdk and it will be easier to install api. Yeah, although I don't like the idea to mess around in /opt. Maybe it would be possible to introduce an android group, so it would be sufficient to add my user to this group? I think it's best to just install it to your home directory without involving the system package manager if you want to use the android package manager for anything. It's not sane to be giving users access to a directory in the PATH of other users, especially root. Many people have a single user system without a true non-administrator account, but packages need to be secure in cases where this isn't true. signature.asc Description: OpenPGP digital signature
[arch-general] KDE update and baloo
After the update to 4.13 - i have baloo sucking up cpu cylces. I did go to the desktop search and add every single file system/directory to the 'dont scan' list. So there should be nothing left to scan. Didn't help - it's still running 2 hours later ... anyone know how to stop this selfish cpu hog? thanks!
Re: [arch-general] Problem with Wifi Connections (Intel Corporation PRO/Wireless 4965 AG Connection)
Hi Sean, I have been having issues connecting to a WiFi network on my laptop lately as well, here's what I did to resolve the issue, which might help in your case too. I was having problems with netctl because it apparently is trying to use dhcpcd in the wrong way, or is trying to take on the role that dhcpcd plays. Maybe someone can correct me if I'm misinterpreting the issue at hand. The way I resolved this issue was to revert back to netctl-1.4-2, run `ip link set wlan0 down' (wlan0 is my wireless interface), and then run `netctl enable wlan0-network' (wlan0-network is my netctl profile). Hopefully this helps, Zack Gold On Thu, Apr 17, 2014 at 3:57 AM, Maximilian Bräutigam m...@xbra.de wrote: Am 16.04.2014 17:27, schrieb Sean Snell: Hi everybody, First, this is on a Lenovo X61 laptop. This is a fresh installation on a new SSD hard drive I recently purchased. I already have a 500GB hard drive with Arch and Windows dual booting on it, and the Wifi still works just fine from the old installation when I reinstalled the old hard drive for testing, but for the life of me I cannot get wifi-menu to connect to any wifi connections on this new installation. The card is a Intel Corporation PRO/Wireless 4965 AG. Per the wireless wikihttps://wiki.archlinux.org/index.php/Wireless_network_configuration#iwlegacy, I've installed linux-firmware so I can use the iwlegacy (driver?). With the help of a friend, we feel the modules are loading correctly, but when we try to actually connect using wifi-menu, journalctl -xn says the following: \\\ - Unit netctl@wls3\x2dNEOPCS\x2dSecured.service has begun starting up. Starting network profile 'wls3-NEOPCS-Secured'... The interface of network profile 'wls3-NEOPCS-Secured' is already up netctl@wls3\x2dNEOPCS\x2dSecured.service: main process exited, code=exited, status=1/FAILURE \ What I've installed in relation to the Wifi: - wpa_supplicant (per the initial insistence of wifi-menu out of the box - linux-firmware (per the wikihttps://wiki.archlinux.org/index.php/Wireless_network_configuration#iwlegacy ) Output of lspci | grep Intel: 00:00.0 Host bridge: Intel Corporation Mobile PM965/GM965/GL960 Memory Controller Hub (rev 0c) 00:02.0 VGA compatible controller: Intel Corporation Mobile GM965/GL960 Integrated Graphics Controller (primary) (rev 0c) 00:02.1 Display controller: Intel Corporation Mobile GM965/GL960 Integrated Graphics Controller (secondary) (rev 0c) 00:19.0 Ethernet controller: Intel Corporation 82566MM Gigabit Network Connection (rev 03) 00:1a.0 USB controller: Intel Corporation 82801H (ICH8 Family) USB UHCI Controller #4 (rev 03) 00:1a.1 USB controller: Intel Corporation 82801H (ICH8 Family) USB UHCI Controller #5 (rev 03) 00:1a.7 USB controller: Intel Corporation 82801H (ICH8 Family) USB2 EHCI Controller #2 (rev 03) 00:1b.0 Audio device: Intel Corporation 82801H (ICH8 Family) HD Audio Controller (rev 03) 00:1c.0 PCI bridge: Intel Corporation 82801H (ICH8 Family) PCI Express Port 1 (rev 03) 00:1c.1 PCI bridge: Intel Corporation 82801H (ICH8 Family) PCI Express Port 2 (rev 03) 00:1d.0 USB controller: Intel Corporation 82801H (ICH8 Family) USB UHCI Controller #1 (rev 03) 00:1d.1 USB controller: Intel Corporation 82801H (ICH8 Family) USB UHCI Controller #2 (rev 03) 00:1d.7 USB controller: Intel Corporation 82801H (ICH8 Family) USB2 EHCI Controller #1 (rev 03) 00:1e.0 PCI bridge: Intel Corporation 82801 Mobile PCI Bridge (rev f3) 00:1f.0 ISA bridge: Intel Corporation 82801HEM (ICH8M-E) LPC Interface Controller (rev 03) 00:1f.2 IDE interface: Intel Corporation 82801HM/HEM (ICH8M/ICH8M-E) SATA Controller [IDE mode] (rev 03) 00:1f.3 SMBus: Intel Corporation 82801H (ICH8 Family) SMBus Controller (rev 03) 03:00.0 Network controller: Intel Corporation PRO/Wireless 4965 AG or AGN [Kedron] Network Connection (rev 61) Output of lsmod | grep -h iw: iwl4965 90075 0 iwlegacy 49001 1 iwl4965 mac80211 490609 2 iwl4965,iwlegacy cfg80211 431038 3 iwl4965,iwlegacy,mac80211 led_class 3547 3 sdhci,iwlegacy,thinkpad_acpi Can anybody help clue me into what I might be overlooking? I've run out of ideas of what else to look into, and while I'm still getting my feet wet with Linux, being a Windows admin by profession, I'm not afraid to dig deeper with a little help! :) Thanks everybody! Sean Hi Sean, to my mind it look promising since the correct modules appear to be loaded. Maybe you could comment an how you try to establish the connection. I usually prefer to use the NetworkManager. https://wiki.archlinux.org/index.php/NetworkManager base package is networkmanager, for Gnome: network-manager-applet for KDE: kdeplasma-applets-plasma-nm I think (but I am unsure) you should stop all netctl services running before trying the following commands. Start the daemon (cases are correct this way): # systemctl start NetworkManager.service If you wanna use cli: List of devices: #
Re: [arch-general] Problem with Wifi Connections (Intel Corporation PRO/Wireless 4965 AG Connection)
Thank you for the replies; I'll be back on the project within the next 2-3 hours, and I'll report back on your suggestions. Sean On Thu, Apr 17, 2014 at 9:31 AM, Zack Gold z...@linux.com wrote: Hi Sean, I have been having issues connecting to a WiFi network on my laptop lately as well, here's what I did to resolve the issue, which might help in your case too. I was having problems with netctl because it apparently is trying to use dhcpcd in the wrong way, or is trying to take on the role that dhcpcd plays. Maybe someone can correct me if I'm misinterpreting the issue at hand. The way I resolved this issue was to revert back to netctl-1.4-2, run `ip link set wlan0 down' (wlan0 is my wireless interface), and then run `netctl enable wlan0-network' (wlan0-network is my netctl profile). Hopefully this helps, Zack Gold On Thu, Apr 17, 2014 at 3:57 AM, Maximilian Bräutigam m...@xbra.de wrote: Am 16.04.2014 17:27, schrieb Sean Snell: Hi everybody, First, this is on a Lenovo X61 laptop. This is a fresh installation on a new SSD hard drive I recently purchased. I already have a 500GB hard drive with Arch and Windows dual booting on it, and the Wifi still works just fine from the old installation when I reinstalled the old hard drive for testing, but for the life of me I cannot get wifi-menu to connect to any wifi connections on this new installation. The card is a Intel Corporation PRO/Wireless 4965 AG. Per the wireless wiki https://wiki.archlinux.org/index.php/Wireless_network_configuration#iwlegacy , I've installed linux-firmware so I can use the iwlegacy (driver?). With the help of a friend, we feel the modules are loading correctly, but when we try to actually connect using wifi-menu, journalctl -xn says the following: \\\ - Unit netctl@wls3\x2dNEOPCS\x2dSecured.service has begun starting up. Starting network profile 'wls3-NEOPCS-Secured'... The interface of network profile 'wls3-NEOPCS-Secured' is already up netctl@wls3\x2dNEOPCS\x2dSecured.service: main process exited, code=exited, status=1/FAILURE \ What I've installed in relation to the Wifi: - wpa_supplicant (per the initial insistence of wifi-menu out of the box - linux-firmware (per the wiki https://wiki.archlinux.org/index.php/Wireless_network_configuration#iwlegacy ) Output of lspci | grep Intel: 00:00.0 Host bridge: Intel Corporation Mobile PM965/GM965/GL960 Memory Controller Hub (rev 0c) 00:02.0 VGA compatible controller: Intel Corporation Mobile GM965/GL960 Integrated Graphics Controller (primary) (rev 0c) 00:02.1 Display controller: Intel Corporation Mobile GM965/GL960 Integrated Graphics Controller (secondary) (rev 0c) 00:19.0 Ethernet controller: Intel Corporation 82566MM Gigabit Network Connection (rev 03) 00:1a.0 USB controller: Intel Corporation 82801H (ICH8 Family) USB UHCI Controller #4 (rev 03) 00:1a.1 USB controller: Intel Corporation 82801H (ICH8 Family) USB UHCI Controller #5 (rev 03) 00:1a.7 USB controller: Intel Corporation 82801H (ICH8 Family) USB2 EHCI Controller #2 (rev 03) 00:1b.0 Audio device: Intel Corporation 82801H (ICH8 Family) HD Audio Controller (rev 03) 00:1c.0 PCI bridge: Intel Corporation 82801H (ICH8 Family) PCI Express Port 1 (rev 03) 00:1c.1 PCI bridge: Intel Corporation 82801H (ICH8 Family) PCI Express Port 2 (rev 03) 00:1d.0 USB controller: Intel Corporation 82801H (ICH8 Family) USB UHCI Controller #1 (rev 03) 00:1d.1 USB controller: Intel Corporation 82801H (ICH8 Family) USB UHCI Controller #2 (rev 03) 00:1d.7 USB controller: Intel Corporation 82801H (ICH8 Family) USB2 EHCI Controller #1 (rev 03) 00:1e.0 PCI bridge: Intel Corporation 82801 Mobile PCI Bridge (rev f3) 00:1f.0 ISA bridge: Intel Corporation 82801HEM (ICH8M-E) LPC Interface Controller (rev 03) 00:1f.2 IDE interface: Intel Corporation 82801HM/HEM (ICH8M/ICH8M-E) SATA Controller [IDE mode] (rev 03) 00:1f.3 SMBus: Intel Corporation 82801H (ICH8 Family) SMBus Controller (rev 03) 03:00.0 Network controller: Intel Corporation PRO/Wireless 4965 AG or AGN [Kedron] Network Connection (rev 61) Output of lsmod | grep -h iw: iwl4965 90075 0 iwlegacy 49001 1 iwl4965 mac80211 490609 2 iwl4965,iwlegacy cfg80211 431038 3 iwl4965,iwlegacy,mac80211 led_class 3547 3 sdhci,iwlegacy,thinkpad_acpi Can anybody help clue me into what I might be overlooking? I've run out of ideas of what else to look into, and while I'm still getting my feet wet with Linux, being a Windows admin by profession, I'm not afraid to dig deeper with a little help! :) Thanks everybody! Sean Hi Sean, to my mind it look promising since the correct modules appear to be loaded. Maybe you could comment an how you try to establish the connection. I usually prefer to use the NetworkManager. https://wiki.archlinux.org/index.php/NetworkManager base package is
Re: [arch-general] KDE update and baloo
On Thursday 17 April 2014 09:11:35 Genes Lists wrote: After the update to 4.13 - i have baloo sucking up cpu cylces. I did go to the desktop search and add every single file system/directory to the 'dont scan' list. So there should be nothing left to scan. Didn't help - it's still running 2 hours later ... anyone know how to stop this selfish cpu hog? I disabled it by editing $HOME/.kde4/share/config/baloofilerc: [Basic Settings] Indexing-Enabled=false 1. https://mailman.archlinux.org/pipermail/arch-general/2014-March/035350.html -- С уважением, Е.Алексеев. Sincerely yours, E.Alekseev. e-mail: darkarca...@mail.ru ICQ: 407-398-235 Jabber: arca...@jabber.ru signature.asc Description: This is a digitally signed message part.
Re: [arch-general] KDE update and baloo
On Thu, Apr 17, 2014 at 3:11 PM, Genes Lists li...@sapience.com wrote: After the update to 4.13 - i have baloo sucking up cpu cylces. I did go to the desktop search and add every single file system/directory to the 'dont scan' list. So there should be nothing left to scan. Didn't help - it's still running 2 hours later ... anyone know how to stop this selfish cpu hog? thanks! Is it with KDE PIM? I remember a discussion in the release ML that baloo was somewhat broken in PIM. I dunno if that was fixed or not.
[arch-general] On-boot delay due to timer units
Hi, Since anacron jobs were replaced with timers, I am seeing a noticeable delay before agetty prompt appears on machines which were unused for some time (due to update/man-db timers starting up simultaneously). TLDR: Anacron inserts a random delay between boot and running the jobs, so is it possible to simulate this behavior by including e.g. OnBootSec=... in the timers at next update? Or is this option incompatible with OnCalendar? Here is the (edited) statistics obtained by grepping /var/log/daemon.log. The disk is actually an Intel X-25 (sata-2) SSD. --- No timers are active (baseline) --- Apr 6: 5.983s (kernel) + 1.947s (userspace) = 7.930s. Apr 6: 5.815s (kernel) + 2.494s (userspace) = 8.310s. Apr 6: 5.692s (kernel) + 1.612s (userspace) = 7.304s. Apr 7: 5.874s (kernel) + 2.561s (userspace) = 8.436s. Apr 9: 5.704s (kernel) + 3.001s (userspace) = 8.706s. Apr 10: 5.612s (kernel) + 2.494s (userspace) = 8.106s. Apr 11: 5.618s (kernel) + 2.908s (userspace) = 8.526s. Apr 12: 5.671s (kernel) + 3.345s (userspace) = 9.016s. --- Timers first run --- Apr 14: 5.464s (kernel) + 46.883s (userspace) = 52.348s. --- Startup with timers --- Apr 15: 5.715s (kernel) + 2.878s (userspace) = 8.593s. Apr 16: Not powered on Apr 17: 6.414s (kernel) + 7.785s (userspace) = 14.200s. $ systemd-analyze blame | head 6.724s man-db.service 1.935s updatedb.service 926ms root-ssh-key-init@0x14d33aba.service 507ms lxc@appserver\x2dx86_64.service 427ms rfkill-unblock@wlan.service 381ms systemd-networkd.service 340ms wlan-powersave@wls1.service 289ms syslog-ng.service 235ms volatile-mail.service 225ms iptables.service Thanks, L. -- Leonid Isaev GPG fingerprints: DA92 034D B4A8 EC51 7EA6 20DF 9291 EE8A 043C B8C4 C0DF 20D0 C075 C3F1 E1BE 775A A7AE F6CB 164B 5A6D signature.asc Description: PGP signature
Re: [arch-general] On-boot delay due to timer units
Am 17.04.2014 20:56, schrieb Leonid Isaev: Hi, Since anacron jobs were replaced with timers, I am seeing a noticeable delay before agetty prompt appears on machines which were unused for some time (due to update/man-db timers starting up simultaneously). TLDR: Anacron inserts a random delay between boot and running the jobs, so is it possible to simulate this behavior by including e.g. OnBootSec=... in the timers at next update? Or is this option incompatible with OnCalendar? OnBootSec would cause the timers to always run on boot, no matter how much time has passed, which is not what we want. I don't think it is a problem that the timers run on boot, but rather that they delay Type=idle units, like agetty. From what the documentation says, there should not be any delay: Behavior of idle is very similar to simple; however, actual execution of the service binary is delayed until all jobs are dispatched. I am confused why get a delay here. I think another solution in systemd would be introducing a holdoff time: Instead of running immediately on boot, the timer should be scheduled for boot+5min. This requires some investigation - sorry, I don't have a quick solution right now. signature.asc Description: OpenPGP digital signature
Re: [arch-general] KDE update and baloo
On 04/17/2014 10:37 AM, Lukas Jirkovsky wrote: Is it with KDE PIM? I remember a discussion in the release ML that baloo was somewhat broken in PIM. I dunno if that was fixed or not. No not using PIM. I killed off all the baloo processes - short term fix.There really should be an off button on this kind of thing.
Re: [arch-general] On-boot delay due to timer units
On Thu, 17 Apr 2014 21:31:07 +0200 Thomas Bächler tho...@archlinux.org wrote: Am 17.04.2014 20:56, schrieb Leonid Isaev: Hi, Since anacron jobs were replaced with timers, I am seeing a noticeable delay before agetty prompt appears on machines which were unused for some time (due to update/man-db timers starting up simultaneously). TLDR: Anacron inserts a random delay between boot and running the jobs, so is it possible to simulate this behavior by including e.g. OnBootSec=... in the timers at next update? Or is this option incompatible with OnCalendar? OnBootSec would cause the timers to always run on boot, no matter how much time has passed, which is not what we want. OK. I don't think it is a problem that the timers run on boot, but rather that they delay Type=idle units, like agetty. From what the documentation says, there should not be any delay: Behavior of idle is very similar to simple; however, actual execution of the service binary is delayed until all jobs are dispatched. I am confused why get a delay here. I think the problem is the disk I/O generated due to e.g. man-db indexing, because I see the hdd light is solid on. So, my guess is that two things can happen: either the login prompt is delayed, or the prompt is shown but the actual login will stall. I think another solution in systemd would be introducing a holdoff time: Instead of running immediately on boot, the timer should be scheduled for boot+5min. You are right -- that's the best way to put it. Except, I'd generate random timeouts (distributed in some interval) for the corresponding services... Thanks, L. -- Leonid Isaev GPG fingerprints: DA92 034D B4A8 EC51 7EA6 20DF 9291 EE8A 043C B8C4 C0DF 20D0 C075 C3F1 E1BE 775A A7AE F6CB 164B 5A6D signature.asc Description: PGP signature
Re: [arch-general] On-boot delay due to timer units
On Thu, Apr 17, 2014 at 9:31 PM, Thomas Bächler tho...@archlinux.org wrote: I don't think it is a problem that the timers run on boot, but rather that they delay Type=idle units, like agetty. From what the documentation says, there should not be any delay: Behavior of idle is very similar to simple; however, actual execution of the service binary is delayed until all jobs are dispatched. I am confused why get a delay here. When the timer fires it adds a start job to the manager. Type=idle services wait for the manager job list (not the transaction job list) to empty. Maybe Type=idle should be changed to trigger when its transaction completes.
Re: [arch-general] Android support in Linux Arch
Hi, Am 17.04.2014 08:53, schrieb Daniel Micay: I think it's best to just install it to your home directory without involving the system package manager if you want to use the android package manager for anything. So you are managing all of this alone and don't use the packages in AUR at all? At least android-udev seems useful? Where do you store your SDK? Have you added something to your PATH? I'm just looking for good practices here, so thanks for any replies! It's not sane to be giving users access to a directory in the PATH of other users, especially root. Yes, I feel very much the same way. Best regards, Karol Babioch signature.asc Description: OpenPGP digital signature
Re: [arch-general] Android support in Linux Arch
On 17/04/14 05:22 PM, Karol Babioch wrote: Hi, Am 17.04.2014 08:53, schrieb Daniel Micay: I think it's best to just install it to your home directory without involving the system package manager if you want to use the android package manager for anything. So you are managing all of this alone and don't use the packages in AUR at all? At least android-udev seems useful? Where do you store your SDK? Have you added something to your PATH? I'm just looking for good practices here, so thanks for any replies! I use the AUR packages without touching the Android package manager at all. If I did plan on installing stuff from there, I would just install the SDK to my home directory and use a meta-package (no contents) to handle the external dependencies. You can add user-specific binary/library directories to PATH and the library search path (a fairly sane usage of LD_LIBRARY_PATH). It's not sane to be giving users access to a directory in the PATH of other users, especially root. Yes, I feel very much the same way. Best regards, Karol Babioch signature.asc Description: OpenPGP digital signature
Re: [arch-general] Android support in Linux Arch
Hi, Am 17.04.2014 23:29, schrieb Daniel Micay: I would just install the SDK to my home directory and use a meta-package (no contents) to handle the external dependencies. Yes, that would probably make sense. Maybe such a package could be uploaded to the AUR. I've installed it to my home directory for now. The only dependency I needed to get it running was swt. I'm not sure why the AUR packages lists all of the lib32 as dependency (for x86_64). I guess I'll find out soon enough. Best regards, Karol Babioch signature.asc Description: OpenPGP digital signature
Re: [arch-general] KDE update and baloo
On Thu, Apr 17, 2014 at 4:33 PM, Genes Lists li...@sapience.com wrote: On 04/17/2014 10:37 AM, Lukas Jirkovsky wrote: Is it with KDE PIM? I remember a discussion in the release ML that baloo was somewhat broken in PIM. I dunno if that was fixed or not. No not using PIM. I killed off all the baloo processes - short term fix.There really should be an off button on this kind of thing. I've a problem with baloo too, but with RAM usage. After a few minutes of KDE with baloo activated and indexing data in some partitions with a few hundred gigabytes of my data (video, music, big source trees of projects, VMs, etc), I experience a continuous growing in RAM usage, sometimes up to consuming my entire 12G RAM and starting paging to swap, driving my system almost unusable. Another times, RAM grows up to 4~5 Gb and stay there. I have no KDE PIM configured (no mail accounts, etc). Weird is that that amount of RAM isn't reported by htop/top/ps for being from any processes. The sum of all processes memory usage stay near the habitual 1-1.5G RAM. Logout and terminating all user processes doesn't free my memory too. The only way to get my memory back is restarting PC. But I'm sure the problem is with baloo or something related, since disabling it the problem goes away. Maybe something related with my btrfs root/home, although mentioned large data partitions are in ext4. I don't want to disable baloo, I liked it's speed in file indexing and searching, but it's unusable now. Anyone know if there's any way to fix it? Or, at least, track down where my memory gone? [maybe kernel memory? kernel caches? unfreed pages?]. Thanks -- /-=| Δ ŋ đ г Σ |=-\ «» ♫♫♫ http://www.google.com/profiles/andre.vmatos
Re: [arch-general] Android support in Linux Arch
Looks like my message was silently dropped by mailman. Lemme retry this: On 2014-04-16 20:49, Hugo Osvaldo Barrera wrote: First of all, thanks for all the efort you're putting into moving these arch tools into the official repos. I've been wanting to see this (and non-bin packages) for ages! :) On 2014-04-17 00:50, Karol Babioch wrote: Hi, Am 17.04.2014 00:38, schrieb Anatol Pomozov: Are there people with Android development background? What exactly do you miss in Arch? The problem I face with the Android situation in Arch is that currently there seems to be no clean (TM) way to install the SDK and related stuff. The android-sdk package from AUR is fine and dandy, but one usually also needs to install a whole bunch of API specific packages through the android tool from the SDK. - This doesn't work for normal users, e.g. you can update the packages using Eclipse, but you need to start /opt/android-sdk/tools/android as root Does this download additional files, or actually replace files the arch package installs? If it's the former, then you can create a user group (eg: android), and make the directory where files are downloaded owned by that group. - Installing any sort of package through the installer mentioned above isn't compatible with the whole idea of package management, because the package manager isn't aware of these files. I ran into conflicts before, which I had to resolve by temporarily removing some components. If we can make arch packages for all the packages available through that installed, that would make it innecesary, though still usable. Something similar happens with npm, gem (when used at a system level), pip, etc: there's a second package manager that can (optionally) be used, but it's a bad idea if you want to keep using arch's. Maybe I'm doing something wrong here, but at least this is what I've experienced throughout the last couple of months. Unfortunately I don't see a good way how this can be improved, as I like the idea of installing only API components that I really need and get instant (!) updates for them directly from the upstream project. If you want the instante updated from upstream, then you'd need to update the arch package instantly ;) This is exactly what happens with some of the above mentioned examples (npm). Anyone familiar with the situation on other distributions? How do they handle all of this? I did a bit of research on this. Ubuntu suggest you download the SDK and install into into your home: https://help.ubuntu.com/community/AndroidSDK (so no useful precedent here). The same applies for Fedora: https://fedoraproject.org/wiki/HOWTO_Setup_Android_Development Gentoo uses the upstream binaries in their packages (ebuild?): https://wiki.gentoo.org/wiki/Android They DO seem to set permissions to 775, and ownership to root:android, so I guess they do something similar to what I suggested above. Finally, Debian doesn't seem to package anythis other than the packages that were mentioned as existing in AUR as source packages, so there's nothing to be leart there. Best regards, Karol Babioch Hope this helps a bit, Cheers, -- Hugo Osvaldo Barrera -- Hugo Osvaldo Barrera A: No, it doesn't make sense. Q: Should I include quotations *after* my reply? pgpipbq8yMXoQ.pgp Description: PGP signature
Re: [arch-general] Android support in Linux Arch
Hi, Am 18.04.2014 00:35, schrieb Hugo Osvaldo Barrera: Does this download additional files, or actually replace files the arch package installs? If it's the former, then you can create a user group (eg: android), and make the directory where files are downloaded owned by that group. Well, it probably depends on what exactly you select to install. When there is a new API level, there will definitely be new files. This can lead to file conflicts, because pacman will complain that the file(s) already exist once the package gets updated. This can be resolved easily, but requires some knowledge about package management. Probably nothing a beginner (either to pacman or to the SDK wants to deal with). I did a bit of research on this. Thanks for that! Ubuntu suggest you download the SDK and install into into your home: https://help.ubuntu.com/community/AndroidSDK (so no useful precedent here). The same applies for Fedora: https://fedoraproject.org/wiki/HOWTO_Setup_Android_Development Personally I like this approach the most. Obviously it has drawbacks in multi-user environments. But it won't lead to conflicts, because pacman doesn't know anything about it and to be quite honest most of us are the only user on a system anyway. However, I kind of like the proposed idea of an empty meta package, that will only trigger the installation of dependencies. Is this something you would be interested in? Best regards, Karol Babioch signature.asc Description: OpenPGP digital signature
Re: [arch-general] Android support in Linux Arch
On 2014-04-18 01:20, Karol Babioch wrote: ...snip... Personally I like this approach the most. Obviously it has drawbacks in multi-user environments. But it won't lead to conflicts, because pacman doesn't know anything about it and to be quite honest most of us are the only user on a system anyway. However, I kind of like the proposed idea of an empty meta package, that will only trigger the installation of dependencies. Is this something you would be interested in? Best regards, Karol Babioch I actually use the meta-package approach to handle dependencies for wine-based and steam-based games, so I wouldn't mind (I hate marking dependencies as explicitly installed, so that's a second reason to do that). I'm curious if those are acceptable in the AUR. -- Hugo Osvaldo Barrera pgpvLQkjBDSlf.pgp Description: PGP signature
[arch-general] system-config-printer: conflicting files
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Hello fellow Archers, I received the following error while upgrading system-config-printer: error: failed to commit transaction (conflicting files) system-config-printer: /usr/share/system-config-printer/debug.pyc exists in filesystem Errors occurred, no packages were upgraded. Apparently, at some point python automatically compiled debug.py into debug.pyc to increase long term performance. However, the new package has debug.pyc presupplied. The solution was to delete debug.pyc, then everything installs cleanly and works fine. Of course, this begs the question: Should a compiled python module be supplied in a package (debug.pyc), or should the local machine interpret/compile/do whatever it wants with a python script (debug.py)? - --Kyle -BEGIN PGP SIGNATURE- Version: GnuPG v2.0.22 (GNU/Linux) Comment: Using GnuPG with Thunderbird - http://www.enigmail.net/ iQIcBAEBAgAGBQJTUH7GAAoJEN5rMzXPJBsQvwkP/jrjwgfai7jUTTFI980qQdaM nu6u/kms6Ys72uJIvxF3a0b94ZprbLe6bjshzCBW0fJuBHrZ+Deb1Xdt2i10b3pV O/JgZzYMsICVqF2MYdn+GZ9T/GYkLpzLWctwBCnHwQKABIw+T5Jk4CwdNLTnd4V5 pmxynMi/A3RzYFsxFQLcNK6a1sk/FhVmwDSAk5WB2feKjf/i614dH8Uf0aJxRcq+ LFg+eFndpEpMp77j0DnxcTwAR6RzdE19NCNf9rfBvctGpyE2nzdYE8bcof66mLoB yWR2jmRC2CYwOxU556I+Bg4TIGy4SQunqkc0hKEG+nQ6/xSOV72CF7ZCFBOKrLsW 8rGPHDNXFFbLeE6kcR+hnEYaKjLTLn+o+BwV9CJSB3l3aCXpoBCJ9kDiEPEfPR6h mZwP77BpBUMTfclTYL/F2mR0NDd4615CWxXd6wHXGCex+WrA4FTyGqGQgUvPa5Bu 9yB/oMyLthgGL9S5CN0N2VcRdPQB4X44DhUieOMqWAO+KVKmJGfohkCRRHy7UOCY JLh8dkA9JBnoeQ9wLFXn7ibSFxJAVfUeddnmsKOaRnVSGVjGuKYbb9IcFQ/xcfBa ghnTIOk6l15Rrzd9pxfoXSDz8eIeC9UWGQN13Q6waD1gOjZVPS2rgmM2vNpYzT5M iyOEOWaj6AVvni9eADE1 =tWUo -END PGP SIGNATURE-
Re: [arch-general] system-config-printer: conflicting files
On 2014-04-17 20:24, Kyle Terrien wrote: Hello fellow Archers, I received the following error while upgrading system-config-printer: error: failed to commit transaction (conflicting files) system-config-printer: /usr/share/system-config-printer/debug.pyc exists in filesystem Errors occurred, no packages were upgraded. Apparently, at some point python automatically compiled debug.py into debug.pyc to increase long term performance. However, the new package has debug.pyc presupplied. The solution was to delete debug.pyc, then everything installs cleanly and works fine. Of course, this begs the question: Should a compiled python module be supplied in a package (debug.pyc), or should the local machine interpret/compile/do whatever it wants with a python script (debug.py)? - --Kyle Yes, they should be included in the package. 1. The user running these scripts usually doesn't have write access to the dir they're in, so more often than not, the .pyc files don't get saved. This makes the load times for the script longer. 2. If it is run as root, a file gets created that pacman knows nothing about. It will leave crap behind when upgrading or removing.
