[SOLVED] Questions about delta mirror/package

My bandwidth is slow and expensive so i use delta package as per this Arch Wiki.
But, for anumber of reasons the delta mirror only provide small number of delta package, and it's geographically far away from where i live so it's kind of slower.
Here's my /etc/pacman.d/mirrorlist:
Server = http://suro.ubaya.ac.id/archlinux/$repo/os/$arch
Server = http://delta.archlinux.fr/$repo/os/$arch
First line is nearest Arch mirror dan second line is delta package mirror.
What i want to achieve with that mirrorlist is Arch download update with delta package if available, if there is no delta package avalaible for that update then Arch will download regular package from the nearest mirror.
Thanx
EDIT: Solved it, well, kind of ...
I end up created a script for update.
#!/bin/sh
sed -i '/suro/s/^Server/#Server/' /etc/pacman.d/mirrorlist
pacman -Syy
sed -i '/suro/s/^#Server/Server/' /etc/pacman.d/mirrorlist
pacman -Su
Last edited by si_kabayan (2014-11-13 05:47:29)

karol wrote:I mean the order in the mirrorlist.
OK.
Lone_Wolf wrote:
The order in mirrorlist is what matters.
pacman will ask the first uncommented mirror , if that has the info required pacman uses that mirror.
If the first uncommented mirror doesn't have the info or doesn't respond, pacman will try the next uncommented mirror
Do you know if delta.archlinux.fr hosting only delta package or mirrorring all Arch package?
Because i just need delta package from this mirror, and regular package must from nearest mirror.

Similar Messages

[SOLVED] Question about Saving Cached Packages

Hey Guys,
I've got a couple of systems running Arch right now, and have a 3rd that I was thinking about setting up this weekend. The second of the 2 I already have is very recent and I haven't cleared out the package cache yet. If I were planning on installing the same applications on the next system, could I just save the package files in /var/cache/pacman to a CD or something and just copy them back to the new system once I have the CORE installed? I figure this would save an hour or so of downloading, which would be nice. If it is possible, is there anything else I need to have aside from the packages themselves (ie. is there like a header file somewhere)?
Thanks!
Swill
Last edited by Mr. Swillis (2008-11-21 06:01:15)

Well - it could work - or it might not. Better solution is to make another repository using your recent packages.
Try to write this to your /etc/pacman.conf:
[name_of_your_own_repo]
Include = /path_to_your_own_repo
I haven't tried it personally, but it should work. If it will prefer packages from other repos (community, extra), then write it somewhere above other repos in /etc/pacman.conf or use pacman -S name_of_your_own_repo/package
You should read about Pacman at Arch-wiki to know more about making your own repository before you will decide to try it.
Either way, write your experience here, please, I'd want to know if it works or not
Last edited by cybermage (2008-11-20 20:19:41)

[SOLVED] Question about CPU temperature

Hi everybody,
I just want to ask a question about "standard" CPU temperatures because my laptop (a Dell Vostro 3700) is permanently hot with the fans on even when I am not doing anything special.
I have CPU freq scaling active with the conservative governor, and I use Gnome Shell 3.2 with an nVidia using the nouveau driver.
Here are "stable" temperatures while writing this post:
# sensors
acpitz-virtual-0
Adapter: Virtual device
temp1: +72.5°C (crit = +103.0°C)
nouveau-pci-0100
Adapter: PCI adapter
temp1: +78.0°C (high = +100.0°C, crit = +110.0°C)
coretemp-isa-0000
Adapter: ISA adapter
Core 0: +73.0°C (high = +95.0°C, crit = +105.0°C)
Core 2: +73.0°C (high = +95.0°C, crit = +105.0°C)
And this is the output of mpstat
# mpstat -P ALL
Linux 3.1.6-1-ARCH (fm) 01/04/2012 _x86_64_ (4 CPU)
10:16:50 PM CPU %usr %nice %sys %iowait %irq %soft %steal %guest %idle
10:16:50 PM all 1.59 0.18 0.54 0.02 0.00 0.02 0.00 0.00 97.65
10:16:50 PM 0 1.88 0.15 0.72 0.02 0.00 0.01 0.00 0.00 97.22
10:16:50 PM 1 1.64 0.21 0.40 0.02 0.00 0.01 0.00 0.00 97.73
10:16:50 PM 2 1.47 0.15 0.67 0.02 0.00 0.06 0.00 0.00 97.64
10:16:50 PM 3 1.36 0.20 0.39 0.02 0.00 0.01 0.00 0.00 98.01
The fan, of course, is full speed (very annoying)
I think that these temperatures are quite high, but I don't know what's really causing all this heating because my laptop is basically idle.
Any suggestion?
Thanks.
P.S.: Running Linux fm 3.1.6-1-ARCH #1 SMP PREEMPT Thu Dec 22 09:11:48 CET 2011 x86_64 Intel(R) Core(TM) i5 CPU M 430 @ 2.27GHz GenuineIntel GNU/Linux
Last edited by fm (2012-01-06 09:32:01)

samuvuo wrote:
fm wrote:Just to wrap up this topic, I "solved" the issue by adding these parameters to the kernel:
Just curious: what are the temperatures now? Sensors output after the "fix" would be nice.
It's in the previous posts:
Before (idle)
# sensors
acpitz-virtual-0
Adapter: Virtual device
temp1: +72.5°C (crit = +103.0°C)
nouveau-pci-0100
Adapter: PCI adapter
temp1: +78.0°C (high = +100.0°C, crit = +110.0°C)
coretemp-isa-0000
Adapter: ISA adapter
Core 0: +73.0°C (high = +95.0°C, crit = +105.0°C)
Core 2: +73.0°C (high = +95.0°C, crit = +105.0°C)
After (idle)
acpitz-virtual-0
Adapter: Virtual device
temp1: +63.5°C (crit = +103.0°C)
nouveau-pci-0100
Adapter: PCI adapter
temp1: +68.0°C (high = +100.0°C, crit = +110.0°C)
coretemp-isa-0000
Adapter: ISA adapter
Core 0: +62.0°C (high = +95.0°C, crit = +105.0°C)
Core 2: +64.0°C (high = +95.0°C, crit = +105.0°C)
After (after playing 8 min youtube video)
acpitz-virtual-0
Adapter: Virtual device
temp1: +74.5°C (crit = +103.0°C)
nouveau-pci-0100
Adapter: PCI adapter
temp1: +78.0°C (high = +100.0°C, crit = +110.0°C)
coretemp-isa-0000
Adapter: ISA adapter
Core 0: +73.0°C (high = +95.0°C, crit = +105.0°C)
Core 2: +75.0°C (high = +95.0°C, crit = +105.0°C)
Fan spinning is better (less noisy) after specifying the kernel parameters.
Thanks
Last edited by fm (2012-01-06 13:43:05)

Questions about the 'filesystem' package

After reading some recent threads and bug reports about the 'filesystem' package, I have the following questions:
1. What is the reason for adding systemd-related users and groups to this package? Why not add them using the install script in the systemd package?
2. Why ship /etc/{shadow,gshadow} files at all? Why not create and manage them by `pwconv`, `grpconv` and such (perhaps by an install script in the shadow package)?
Last edited by ackalker (2014-06-27 17:12:52)

I didn't mean the dev ML, arch-general will do just fine. arch-dev-public is read-only fro non-devs and non-TUs.

[SOLVED] Question about resolv.conf

I successfully configured my Arch box with a static IP by following the instructions on the Arch wiki page here but had a couple of questions about configuring /etc/resolv.conf:
1) What is the proper way to obtain the domain name that should be inserted into resolv.conf?
2) What exactly is this domain name used for? I read the man page but still don't really understand it. Why do I need a domain name if I already have a working set of nameservers to resolve IP addresses?
EDIT: @ewaller Yes I edited the OP because I realized that my original problem was that I just hadn't added a domain name in my resolv.conf file. I edited my original post because 1) the instructions are already posted clearly on Arch Wiki so my post really wouldn't be beneficial and 2) so I could ask my real question without making other people waste their time reading through my other nonsense
Last edited by choogi (2010-09-09 04:06:45)

yejun wrote:You can just use public dns like 8.8.8.8 or 4.2.2.1. There's no other way to get dnsserver besides dhcp.
Yes, I know how to get a nameserver, but I'm asking about the domain name that goes in resolv.conf. The Arch wiki tutorial for setting up static IP uses the following resolv.conf as an example:
nameserver 61.23.173.5
nameserver 61.95.849.8
search example.com
My question has to do with what the "search example.com" does. I don't have a domain name associated with my machine, but if I remove that line, then I get "web page unavailable" when I try to load any web site, which suggests that I'm not connecting to any nameservers right?

[SOLVED] Question about GCC error messages.

I'm currently taking a course in C programming, and I have a question about the error messages returned by GCC.
When ever i try to compile a program containing syntax errors, i get cryptic and not very helpfull error messages.
If i try to compile:
#include <stdio.h>
int main(void)
int i;
i = 0
return 0;
I get following from GCC:
error.c: In function â:
error.c:8:2: error: expected â before â
But shouldn't the message be something like:
error.c: In function 'main':
error.c:8:2: error: expected ; before 'return'
or am I mistaken?
/AcId
Last edited by AcId (2010-10-13 17:48:19)

Woah. That is very strange. I've never seen that before.
I assume you're using Arch Linux. What text editor did you use to write the code? Which terminal application are you using?
...those questions are probably unrelated to the problem, but we might as well get them out of the way.

[SOLVED] Question about package updates and ABS

Hello,
I have manually compile my kernel in order to apply a patch my laptop needs to suspend. So I use abs and customize the pkgbuild. A few days ago when the update for kernel26-2.6.33.3-2 came out, I thought I'd just refresh abs, rebuild, and then I'd be up to date. But instead now pacman -Syu tells me that there is still an update for the kernel.
Does building from ABS not get you the latest package update?
Also, what exactly is the package update compared to a software update? I'm guessing that it is an arch only update to the package but the software itself is the same version. Or am I totally off?
Thanks for the help!
Last edited by Fingel (2010-05-12 03:42:38)

When building via ABS, the package installed gets its version from the PKGBUILD ($pkgver-$pkgrel). Pacman just looks at available version vs installed version to tell if you should upgrade.
Your abs tree was probably just not up to date; just because pacman -Syu says your mirror has a package, doesn't mean building via abs _at that moment_ will get that version.
Be sure to run `abs` to update your tree, then when you're editing the PKGBUILD, check which version it's building. In some cases, you can simply increment those numbers right in the PKGBUILD before building as source files are often fetched based on those variables, etc. -- but I don't know if I'd recommend that when building something like the kernel.
Last edited by brisbin33 (2010-05-11 17:08:12)

[Solved] Questions about Arch on T420s

Hello everybody,
I'm relatively new to Arch and switched from Ubuntu last year. Since then I've learned a lot about Linux in general due to the great documentation provided by the Arch community.
I've managed to install Arch on my Thinkpad T420s with an UEFI Setup using Archboot. But still there are open questions.
My first question is about the UEFI System partition and grub2. Should the UEFI partition be mounted to the filesystem? What about updating Grub2 (eg for setting boot parameters) when the UEFI System partition is not mounted to the filesystem. Up to now I've set boot parameters by manually mounting the UEFI partition and editing the Grub2's menu.lst equivalent. I think this is not optimal.
My second question is about some graphical artifacts in Gnome 3. The T420s uses Intel HD 3000 Graphics. Sometimes icons in nautilus disapear or get replaced by black planes when hovering over them with the mouse cursor. Similar black planes replacing icons can be observed sometimes in Evolution or Rhythmbox. Furthermore gnome-terminal becomes transparent sometimes when I'm scrolling or switching between powertop2 tabs. Anyone else observed similar issues on their T420s or on other models with an Intel HD 3000? I use several options for the i915 driver but none of them seem to be related to the graphical artifacts. Furthermore I've disabled dVT in the BIOS.
options i915 modeset=1
options i915 i915_enable_rc6=1
options i915 i915_enable_fbc=1
options i915 lvds_downclock=1
options i915 semaphores=1
My last question is about the USB ports on my T420s. I've 3 USB ports, one on the left side (USB2.0) and two (USB3.0 and another USB2.0) on the back side. When I plug in a HD on the left side everything works fine and the filesystem is automatically mounted. When I plug in the same HD on the ports of the back side the HD isn't even recognized. Interestingly when i boot with the HD plugged into the back side ports everything works fine. Can you point me any directions how to solve this problem?
Hopefully sombody using Arch on a T420s has already solved these issues and could share the solutions. Otherwise I would be grateful for any suggestions.
Sincerely yours,
Tobsen
Last edited by tobsen (2012-02-09 17:15:27)

Hi Tobsen, I have a T420 also. Perhaps my experiences will be some help?
Regarding the USB ports: I installed Arch on my T420 with the Kernal 3.0.3 snapshot and all the USB ports worked out of the box. That sounds like a problem with udev to me, though. Have you checked the wiki article on this?
Regarding the graphical artifacts: For me, when a popup window appears from hovering the mouse over an icon, the window is sometimes blank grey. It's totally inconsistent, and I can fix it by just hovering over it again, so I haven't really bothered to examine the source of the problem. Could this be related to your problem with the GPU? I figured it was a problem with the windows manager I'm using (Enlightenment e17), which is unstable.
Regarding UEFI: I don't use GRUB2, so this may not be good advice, but I personally feel like you should avoid auto-mounting partitions you don't always need. It slows down boot, and if the partition is mounted, user error could damage it. I say you should keep doing what you're doing.

[SOLVED] Question about bash

The default .bashrc file looks like this. When installing rvm I noticed that it is not sourcing scripts under /etc/profile.d.
# ~/.bashrc
# If not running interactively, don't do anything
[[ $- != *i* ]] && return
alias ls='ls --color=auto'
PS1='[\u@\h \W]\$ '
I just solved my issue by adding the following code to .bashrc. I was just wondering if this should be done by default perhaps in the package that installs /etc/skel/.bashrc?
if [ -f /etc/profile ]; then
. /etc/profile
fi
Last edited by MikeW (2014-11-12 05:28:25)

Seems that it is an rvm issue. After a bit of searching for rvm problems I found this post that gave me the rvm specific solution. This is probably the preferred fix since its just rvm that's the issue here.
http://askubuntu.com/questions/40287/et … ng-sourced
The key is serveral posts down...
Installing RVM as multiuser requires the user to run this command:
echo '[[ -s "/etc/profile.d/rvm.sh" ]] && . "/etc/profile.d/rvm.sh" # Load RVM function' >> ~/.bashrc
So I can remove the /etc/profile bit now. Thanks for stearing me in the right direction to fix this.

*Solved* Question about ATI Drivers

Ok, I'm just finished installing, and I know most distro's, ask you to remove the open source drivers, before adding in FGLRX drivers. Looking at the wiki, I didn't noticed if this needed to be done, and after going through the install, I rebooted. It went as planned until kdm went to load. The screen was unreadable. I used the cd media to mount root, and removed catalyst, and re-installed the opensource drivers. All is well atm, so I'm trying to figure out what I missed. My question atm though is during the FGLRX install, it ask's if you want to removed the opensource drivers due to conflict. If I hit yes, will pacman it automatically remove the required drivers, or do I need to manually remove the Fx86 drivers on my own?
I do apologize for not having a /etc/X11/xorg.conf report, as it slipped my mind while fixing everything. I'll try to make a better habbit of getting it made. I did do a aticonig --inital before the reboot, so i know it won't that.
If I missed anything on the wiki, please let me know, I'll look again at it after dinner.
Thank you all for your time in the matter.
Last edited by bigcountryva (2012-09-30 01:01:49)

Silex89 wrote:
Hi there
First off, which installation method are you following? (ABS using the AUR package, the unofficial repository or the AMD installer). Depending on which method you use, you'll have (or not) a resolved dependency. For instance, you have to remove both the open source driver and your former xorg.conf file (safer is to back it up) to be able to install Catalyst with the AMD's site installer.
If you use AUR (with a tool like Yaourt for example) or the repo method, pacman/yaourt should do it for you, but in a lot of cases it doesn't (I switched to Catalyst having my OpenBox session built already, so when I tried to remove the open source driver it complained about all of the apps that I had installed for removing a dependency like libgl or mesa) and you'll have to do a pacman -Rdd xf86-video-ati to make it work
I'm wondering if you remembered to add the mkinitcpio hook (which is needed to make Catalyst work).
What I would suggest is to re-read the wiki page for Catalyst, see if your card is still supported by Catalyst, maybe your card needs Catalyst-Legacy...
Best of Luck with that
I figured it out, and got it working. I had a typo adding the module for flgxr driver. Fixed it and a soft boot later, all is well. I also deicided to make it so I have to "startx" into my gui, incase this happens again, I won't need the installation media. Thanks for the input.

Question about Flash iPhone packager

So since Apple now allows it, Adobe picked up development for it again I heard. My question is, additionally to the possibility of creating iPhone apps from scratch I heard there is a packager that allows you to wrap up Actionscript 3 projects as an iPhone app. I didn't get too many infos about it on google since its mostly the same kind of news turning up.
Can you use a previously made Actionscript 3 game or do you have to start a new one and make it with the iPhone in mind? If you can use a previously made project how do you customize it for the touch controls? Does the packager change it into a file useable in Flash again to add these features? I'm totally clueless about the possibilities and it would be great if someone could tell me.
Is it really that I can make an already made as3 game iPhone compatible as long as I have an Apple developer subscription? Or is it limited to simple apps?
Would love for someone to elaborate this. Thanks

http://help.adobe.com/en_US/as3/mobile/flashplatform_optimizing_content.pdf
http://download.macromedia.com/pub/labs/packagerforiphone/packagerforiphone_devguide.pdf
http://forums.adobe.com/community/labs/packagerforiphone?view=discussions&start=0
http://floatlearning.com/2010/10/key-differences-in-air-for-android-and-ios/

[SOLVED]Questions about e17-svn

I installed e17-svn and e17-extra-svn from commouty repo,it's beautiful...and buggy.It creates several desktop icons for my partitions,i deleted that from Desktop and .e/e/fileman/favorites folder,but they appear again the next time i login.Firefox and thunar freeze many times as well,and the most unacceptable thing is that i can't logout,every time i try logout,the e17 hangs there forever and i have to restart x myself.I use kdm4 as the login manager,and i try slim as well,the problem remain.Anybody knows how to solve those problems?Thanks.
Last edited by ashunter (2008-08-24 22:15:34)

WTF, Something weird going on here. It used to be #pacman -S e-cvs and you got a perfectly useable e17 desktop. I just tried to #pacman -S e17-svn and it says no such package or group so there must be some changes goin' on as a #pacman -S e-cvs gets you nowhere now. Be patient, though--I say to myself
I'll bet pressh is on the job! Hmmm... I don't recall seeing any announcement about a big change in getting enlightenment in arch. Maybe I missed something?
Edit: I guess I did miss something. Searching the forums brings me this: http://wiki.archlinux.org/index.php/E17 from a post in 2008-07-03 http://bbs.archlinux.org/viewtopic.php?id=51202
I'm beginning to feel like Rip Van Winkle
Edit: I just did a #pacman -Syu on my x86_64 machine and it asked if I wanted to replace e-cvs with e-svn stuff so I did and sofar no probs. Sorry ashunter, you might have other issues. I use gdm on my desktop and slim on my laptop and no probs logging in or shutting down here.
Last edited by bgc1954 (2008-08-23 16:26:10)

Question about [org.openide] Package

Hi all,
I want to add to my GUI a popup ColorPicker
I found an example that imports org.openide.windows.WindowManager
, but NetBeans 5 that I am using shown to
me that org.openide.windows package
does not exist.
What I should to solve this problem?
if you have another examples about
popup color picker could you show
me that?
Thanks.

The example that I have mentioned located in the follwoing address:
http://www.netbeans.org/kb/articles/desktopsampler/DesktopSamplerToolbar.html
And
I prefer the popup color picker rather than using JColorChooser, You can see here JColorChooser gives you a lots of tools. I just need from
user to select a specific color from the color picker and the JTextField
value will be changed according to the selection.
So, I need first to fix the import org.openide.windows package first.

[Solved] Questions about netcfg

Hello,
I just set up a very basic Arch install and to keep in this spirit, I decided to go with netcfg. I got it all working but I do have some questions, especially about it I need certain packages.
1: At home we have a wifi network that is secured with WPA. I take it, according to the wiki, that I have to use wpa_actiond for this? Or is this just for roaming? (as this part of the wiki suggests "required for automatic/roaming wireless connections".)
2: I'm using wifi-select as well for when I'm on the go. I have added this information as suggested per the wiki to /etc/rc.conf, to automatically connect to my home network? Is this redundant when using wifi-select and/or will this interfere with wifi-select when I want to connect to another network?
I'm a little struggling with the English to get my point across in these questions but I think you'll understand. If not, I'll try to elaborate a little more. Thanks in advance for your help!
Last edited by Unia (2012-01-31 21:29:46)

Raynman wrote:1: That (the quote) is not a suggestion, it's a statement. So your assumption is not correct and I don't understand why you say 'according to the wiki'. wpa_supplicant handles the wpa stuff -- see below for more about wpa_actiond.
I say so, because as I said English is not my first language.
Raynman wrote:2: As the wiki says, net-auto-wireless is useful for roaming, moving between a number of known networks (for example, I have profiles for home, university and public wifi on the train). It also says that this is the part of netcfg that requires wpa_actiond. If you usually use your home network, you might opt for the simpler net-profiles and rely on wifi-select for when you occasionally need to connect to some other network; then you can drop wpa_actiond.
Thanks, I get it! I'll mess around with it tomorrow.

A question about CLASSPATH and Package

The configuration of my computer is:
Windows98, C:\j2sdk1.4.0
Which in Autoexec.bat are included like this:
path=C:\j2sdk1.4.0\bin
set CLASSPATH=C:\j2sdk1.4.0\lib
Then I make a new directory named t in C:\j2sdk1.4.0\lib, where there is a compiled file a.class, source code is:
package t;
public class a
And another java file was in D:\t, named b.java, code like this:
import t.*;
class b extends a
public static void main( String[] args )
System.out.println("OK");
Then I compiled like this: javac b.java, it was wrong; but when I compiled like this: javac -CLASSPATH C:\j2sdk1.4.0\lib b.java, It was OK!!! Why?? And then when I execute like this: java b, an error message was prompted(Can not find class t/a)!!! How can I do?

First one general comment: I would not advise putting your own classes into C:\j2sdk1.4.0\lib since this suggests that the classes under this directory belong to the Java distribution. So my advice: create your own directory somewhere like D:\user\Classes or whatever.
The directory (-ies) you specify in CLASSPATH are root entries telling the JVM where to start looking for classes. Why root entries? Well, the package structure is reflected in the directory structure relative to these roots. So a class of package t belongs into the subdirectory t of one of the directories specified in your CLASSPATH. This would be the case of your class a - by the way, the convention says that class names are capitalized. Since class b, on the other hand, is not in any package, it belongs irirectly into one of the directories of your CLASSPATH.
Whereas the compiler is fairly forgiving about the location of your source files, the JVM is not. Invoking a class of some package requires you to provide the fully-qualified name to the java command. So if you wanted to run class a, you would have to invoke "java t.a". Starting class b, simply requires a call of "java b". But make sure the class file of class a is located in the subdirectory t of one of the directories specified in your CLASSPATH.

[SOLVED] Questions about delta mirror/package

Similar Messages

Maybe you are looking for