Re: The new skype-4.3.0.37 is out. Build your own rpm for Fedora 20
by Sudhir Khanger
On Fri, Jun 20, 2014 at 9:09 PM, Paul Cartwright <pbcartwright(a)gmail.com> wrote:
> What is wrong with the official RPM? sudo yum -y install libXv.i686
> libXScrnSaver.i686 qt.i686 qt-x11.i686 pulseaudio-libs.i686
> pulseaudio-libs-glib2.i686 alsa-plugins-pulseaudio.i686 qtwebkit.i686
> sudo yum install skype-<version>-fedora.i586.rpm works just fine.
>
> No package skype-4.3.0.37-1.fc20.i686.rpm available.
> No packages marked for update
> [root@pauls-server backups]# yum update skype-4.3.0.37-1.fc20.i586.rpm
> Loaded plugins: langpacks
> No Match for argument: skype-4.3.0.37-1.fc20.i586.rpm
> No package skype-4.3.0.37-1.fc20.i586.rpm available.
> No packages marked for update
> [root@pauls-server backups]# yum update skype-4.3.0.37-1.fedora.i586.rpm
> Loaded plugins: langpacks
> No Match for argument: skype-4.3.0.37-1.fedora.i586.rpm
> No package skype-4.3.0.37-1.fedora.i586.rpm available.
> No packages marked for update
> [root@pauls-server backups]# yum update skype-4.3.0.37-1.fedora.i686.rpm
> Loaded plugins: langpacks
> No Match for argument: skype-4.3.0.37-1.fedora.i686.rpm
> No package skype-4.3.0.37-1.fedora.i686.rpm available.
> No packages marked for update
>
>
> --
> Paul Cartwright
> Registered Linux User #367800 and new counter #561587
>
What are you doing? Skype is a closed source package. What made you
think it might be available in Fedora repository.
You are suppose to download the Skype package from official source skype.com.
64 bit system - Install both dependencies and skype package.
32 bit system - Just install the skype package.
For those wanting hassle free updates I highly suggest negativo's repository[1].
[1] http://negativo17.org/skype-and-skype-pidgin-plugin/
--
Regards,
Sudhir Khanger.
sudhirkhanger.com
https://github.com/donniezazen
9 years, 10 months
yum update from 3.14.4 looks ok, but still stuck 3.14.4-200.fc20.x86_64
by Jackson Byers
yum update from 3.14.4 looks ok
in /boot
vmlinuz-3.14.3-200.fc20.x86_64
vmlinuz-3.14.4-200.fc20.x86_64
vmlinuz-3.14.7-200.fc20.x86_64
I can't seem to get 3.14.7 to 'take'
after reboot still in 3.14.4
exact command used for update:
yum update && grub2-mkconfig -o /boot/grub2/grub.cfg
followed by a reboot,
still get:
$ uname -r
3.14.4-200.fc20.x86_64
9 years, 10 months
SMB Browser Election Thread (Was: Case for an occasional system refresh or clean install)
by Temlakos
To Tim in Australia:
That must be it. The machine I was retiring (see the other thread) was
indeed the Master Browser. The replacement machine and the laptop both
accepted it in that role.
And after I retired the machine, neither of the other machines could browse.
After awhile, they would see the Windows machine. Which does not run all
the time, by the way, but I will address that in another thread.
Anyway: after several hours of running the replacement machine and the
laptop, and correcting the configuration file and service-start
instructions on both, they seem to have settled on a regular Master
Browser. Hosts come up very quickly, without delay.
With that said: you mentioned installing Network File System in a
Windows environment. (Mine's Windows 8.1.) I'd like to know how to get
NFS to run between two Linux boxes. That would be a start. Then I could
think about installing an NFS service and configurator on the Windows
machine.
Temlakos
9 years, 10 months
Video capture in Windows and Linux
by Temlakos
Tim in Australia:
You asked, in another thread:
> I'd be interested to know what you use for that (software and hardware).
> I've yet to find anything that isn't painful, or actually works.
The current hardware:
Lenovo IdeaCenter K450. 12 GB RAM, Intel Core i5, clock speed > 3 GHz.
Has three clock-speed settings and a "hardware switch" to set them. I
run it in Turbo mode, described as suitable for gaming.
Hauppauge Computer Works' HD-PVR2, the Gaming Edition. Set for a Sony
Playstation. Accepts component video from a set-top box. (They can't
encrypt component video, but they can and do encrypt HDMI.)
Hauppauge Video Capture application and service. The service listens to
a physical button on the capture box. Push the button, and the capture
will begin at once.
The product: a video using the H.264 video and MP4 audio codecs.
Sampling rate: set at 14 Kbps and constant. Uses the Transfer Stream
(.ts) container.
I then use Nero's Platinum (version 14) software for authoring of DVD
and Blu-Ray disks. To produce the latter, I replaced the existing
optical drive with a Blu-Ray burner from LG. (I forget the model number
just now. It's rated at 14-spin.)
Hauppauge Computer Works is the clear winner for video capture devices.
And they are the only company that builds a capture device for American
high-definition television (1920 x 1080, interlaced).
Level of satisfaction: fair. Every now and again, a captured video will
not transcode. It will hang at a certain scene--always the same scene if
I try to abort and restart, so nothing will serve except to capture it
all over again. (Which is why I always capture a DVR recording, or at
least have a DVR running to back me up in case the live capture fails.)
I have found that nothing less than two hours of total down time
(physical power-off shutdown) will suffice to cool down the processor
before I can make another capture. (If I try to "just be patient" and
let the authoring tool complete the transcode, it will transcode all
right--but the optical disk it produces will be unplayable--a
"coaster.") And even then--well, just now I am waiting to try for the
third time to capture a recording. And I have enough coasters to set a
table for an IT banquet.
Now anyone here can jump in to tell me whether the MythTV project has
anything going with the Hauppauge HD-PVR2. And how soon Linux will be
able to "just play" a Blu-Ray disk, either commercial or one of my
home-burn jobs.
Temlakos
9 years, 10 months
Why won't smbd and nmbd start automatically on system start?
by Temlakos
Everyone:
I have two computers, one desktop and one laptop. I did clean installs
on both. In the process, I retired a third computer that had been
running samba without incident for years.
Today I can get samba running on both machines. But: I have to execute
two commands:
$ sudo /sbin/smbd -D
$ sudo /sbin/nmbd -D
by hand, in a terminal (Konsole), every time I start or restart either
computer.
The system-config-services app fails to note that smbd or nmbd are even
available for starting.
How do I get those two daemons to start automatically, so that I don't
have to type those two commands every time? Because until I do, I have
no file-sharing capability. (I have yet another computer on my network:
a dedicated Windows box that I use for video capturing. I've said on
other threads I am not satisfied with the video-capture and DVD
authoring support Linux provides, and note that Linux does not support
Blu-ray, in playback or especially in burning. So samba is a must for me.)
Temlakos
9 years, 10 months
Case for an occasional system refresh or clean install
by Temlakos
Everyone:
No complaints here. Just some insights from some recent experiences.
Bottom line: everyone who administers a Fedora system, should do a
"clean install" or an effective system refresh (reinstalling the OS and
all apps on a system having a separate /home partition) at least, I
would say, once for every three new versions of Fedora. The "fedup" app
lets too many minor faults accumulate in a system over time. Result: you
are cheating yourself and your users of a good user experience.
Long ago I did a yum upgrade from F14 to F17. At first everything went
swimmingly. And then I got the problem some of you might remember my
complaining about: after two hours of operation, or maybe four or five
or six hours if I were especially blessed (or as quickly as five minutes
if I were under a special curse), new windows and dialog boxes would
open as an unrelieved black. (Some KDE pop-up menus and notifications
would pop up as a transparent outline.) I filed bug after bug after bug
with Bugzilla. I re-filed with every new upgrade, which I accomplished
with fedup. From F17->F18, F18->F19, and F19->F20, the problem
persisted. Frankly I'm surprised the PTB's didn't CLOSE the bug as WONTFIX.
Two weeks ago I acquired a new, replacement machine for one of the two
mini-towers on myh network. At the same time, the mouse on my F20
machine went wild, opening and closing windows and pushing buttons
without my conscious input, so I couldn't stop it. Finally the mouse
just quit. I figured out what happened: the mouse cable had frayed.
Happily, with the new shipment, I got a USB mouse and a USB keyboard
that I now can keep as spares.
But I decided to put F20 on another machine I had intended to replace.
(It was a video-capture machine, and I needed a faster processor with
more memory to do it properly.) I figured, why not go from a Pentium
Dual-Core with 3 GB RAM and a 500-GB HDD to an Intel Core 2 Quattro with
8 MB RAM and a 1TB HDD?
This involved downloading and burning an F20 KDE live spin, inserting it
into the intended replacement machine, and doing a complete system
format--reclaim the drive space, put in the new automatic partition (50
GB for /, 8 GB of swap, 500-MB physical boot partition, and all the
rest, 790 GB, for /home), install F20 from the Internet, and
painstakingly replicate my old environment--or at least those parts of
it I wanted. Fresh install of samba, Firefox, Thunderbird, gnucash, and
any other application I used regularly.
I learned later that the old machine had been providing a crutch for
samba connectivity to other machines on the network. When I retired the
old machine, I lost that. Temporarily. I had to relearn everything I'd
learned and forgotten about configuring samba by directly editing the
smb.conf file. But now I have two machines--a desktop and a laptop--that
can establish samba connectivity independently.
But that black rectangle problem I complained about? That is now solved!
And something else: dialog buttons have a new, "more modern" look. This
should have come with the F19->F20 fedup upgrade. But it did not. It
came with a clean install on a brand-new physical box, to which I have
migrated all my "stuff."
I can only conclude that the yum upgrade introduced the first of several
cumulative errors. Every upgrade since then, introduced another one, and
another, and another. These errors pile up. They won't cause a
catastrophic failure, but they will create several annoying problems you
might never solve.
To the development team that gave us fedup, listen up! Your application
is not taking care of all these little things.
With /home partitioned off separately (as an LVM partition), I assume I
can install F21 "cleanly" in the / partition without risking losing the
contents of /home or my access to them. But I still have to write down
every modification I made to a configuration file--or else keep copies
of those files in /home. They would include, I also assume, things like
the files holding the UIDs, GIDs, and passwords for Users and Groups.
That's the only reason I don't plan to do a clean install every time.
But I wouldn't mind doing it every other time, or every third time. To
make sure these cumulative errors don't build up again.
Again, I offer this to provide some insight into problems that we
"amateur system administrators" sometimes have.
Temlakos
9 years, 10 months
Routing table suddenly limited
by Temlakos
Everyone:
Have you ever seen this result from the command route -n?
Kernel IP routing table
Destination Gateway Genmask Flags Metric Ref
Use Iface
0.0.0.0 192.168.1.1 0.0.0.0 UG 1024
0 0 wlp3s0
192.168.1.0 0.0.0.0 255.255.255.0 U 0 0 0 wlp3s0
The "wlp3s0" is a wireless interface.
But I have another computer that uses a regular Ethernet interface. The
result, except for a different interface name, is the same.
As a result, Samba suddenly won't work for me anymore.
The only difference is, I permanently retired an old computer from my
home network. That computer had been running Fedora 20, and I had made
updates since Fedora 12. Now, with that computer no longer part of the
network, neither of the two Linux boxes can see the other. But either of
them can see a Windows physical box, so long as that box is turned on.
Temlakos
9 years, 10 months
How to create a single 4GB large ramdisk?
by Clemens Eisserer
Hi,
I need a single ramdisk which is a real block device (unlike tmpfs),
so I found the ramdisk_size kernel boot parameter.
However, after invoking "MAKEDEV ram" I get hundreds of 16mb large
ramdisks, instead of a single 4GB large:
dd if=/dev/zero of=/dev/ram0 bs=1024k count=2048
dd: error writing ‘/dev/ram0’: No space left on device
17+0 records in
16+0 records out
16777216 bytes (17 MB) copied, 0.032487 s, 516 MB/s
Any idea how to get a single ramdisk instead, I am using Fedora20 on x86_64 btw.
Thank you in advance, Clemens
9 years, 10 months
5tFTw: F21 on Track, FESCo Membership Change, More DNF vs. Yum, Reading VM Logs, and Wiki Infoboxes (2014-06-17)
by Matthew Miller
Reposted from <http://fedoramagazine.org/5tftw-2014-06-17/>.
Fedora is a big project, and it’s hard to follow it all. This series
highlights interesting happenings in five different areas every week.
It isn’t comprehensive news coverage — just quick summaries with links
to each. Here are the five things for June 17th, 2014:
Fedora 21 Is Not Delayed
------------------------
It might seem a bit funny to report on something that’s *not*
happening, but this got picked up by some of the tech press so I
thought it best to clarify the actual situation. Short version: we’re
*not* currently planning any delay from the established target of
October 14th, 2014.
Longer version: the Fedora Server Working Group is writing new code
which will provide a programmatic interface (an API) for provisioning
servers with certain Server Roles. This doesn’t compete with existing
config management tools like Puppet, Chef, or Ansible — it provides a
framework that they (or the new Cockpit web GUI) can talk to. That’s
taking a little bit longer than planned, especially because new input
from the CentOS Simple Linux Server SIG was incorporated. However, at
last week’s meeting, FESCo (the Fedora Engineering Steering Committee,
which manages the overall schedule) decided to allow *just* this Server
Role Deployment Framework]framework] to land a little late in the
Fedora Alpha schedule, but with no delay to the rest of the project or
change to the overall F21 schedule.
* http://fedoraproject.org/wiki/Server
* http://fedoraserver-wgblog.rhcloud.com/primary-goals-for-fedora-server-in...
* https://fedoraproject.org/wiki/Changes/FrameworkForServerRoleDeployment
* http://cockpit-project.org/
* http://wiki.centos.org/SpecialInterestGroup/SLS
* http://meetbot.fedoraproject.org/fedora-meeting/2014-06-11/fesco.2014-06-...
Bill Nottingham Steps Down From FESCo; Kyle McMartin Steps Up
-------------------------------------------------------------
Long-time and almost omnipresent Fedora contributor Bill Nottingham
has started a new day job at open source config-management company
Ansible, and resigned his Fedora Engineering Steering Committee
seat in order to be able to focus on that. Our election policy
states that open seats are offered to previous runners-up, which
means Kyle McMartin is our newest FESCo member (although not, by
any means new to Fedora). Thank you Bill for years of work on
Fedora in FESCo and elsewhere, and welcome Kyle!
P.S. We use Ansible quite heavily in Fedora Infrastructure. Also:
Former Fedora Project Leader Greg DeKoenigsberg is also joining
Ansible, which already boasts quite the slate of former Red Hatters and
Fedora contributors. I’m looking forward to further collaboration
across our open source projects.
* http://fedoraproject.org/wiki/User:Notting
* https://twitter.com/ansible/status/478518891665432576
* http://fedoraproject.org/wiki/Development/SteeringCommittee/Nominations#K...
* http://www.slideshare.net/AdityaPatawari/ansible-33223245
* http://www.ansible.com/management-team
DNF as Yum Replacement, Continued
---------------------------------
Last week, I wrote about a proposal to replace Fedora’s Yum package
manager with DNF. That proposal — targeting next year’s Fedora 22,
_not_ Fedora 21 this fall — is now officially on the table for
discussion. That discussion continues on the Fedora devel mailing list.
Although there are several long threads, it looks like people are
generally open to the idea and advantages of the new code, but there
are differing opinions on whether DNF should keep that name or become
“Yum 4” — or even take a new generic name like `package`. Feel free to
join in, although please remember to be constructive and to follow the
Fedora Project Code of Conduct (which, basically, boils down to “be
excellent to each other”.
* http://fedoramagazine.org/5tftw-2014-06-10/
* http://yum.baseurl.org/
* http://dnf.baseurl.org/
* https://fedorahosted.org/fesco/ticket/1312
* https://lists.fedoraproject.org/pipermail/devel/2014-June/thread.html
* https://fedoraproject.org/code-of-conduct
Coming Soon to libguestfs: Smart Log Reading
--------------------------------------------
The systemd journal brings a lot of useful features, including
structured log formats and reliable connection of messages to the
services they came from. But, its binary log format can be less
convenient when you’re not working *on* the system where the logs
reside.
This is often the case when using libguestfs, an almost-magically-
powerful tool for working with virtual machine disk images. Developer
Richard W.M. Jones blogs about a new feature: virt-log. This command
automatically does the right thing with several different guest system
log possibilities, including understanding the journald format, knowing
that Debian-based systems use `/var/log/syslog` instead of our
`/var/log/messages`, and in the future will even read the Microsoft
Windows Event Log.
* http://0pointer.de/blog/projects/journalctl.html
* http://libguestfs.org/
* http://rwmj.wordpress.com/2014/06/14/new-in-libguestfs-virt-log/
Add an Infobox to Your Fedora Wiki User Page
--------------------------------------------
Our wiki serves many purposes (maybe too many!), but one of the the
most useful is providing simple autobiographical and contact
information for contributors. A neat little thing I stumbled across
this week is the Infobox template, defined and documented at
<https://fedoraproject.org/wiki/Template:Infobox_user>. Follow these
instructions and add the template to your page for a clean, consistent
presentation of your Fedora-relevant details.
As a bonus, display your Fedora Badges with `{{ #fedorabadges: fasname }}`
* https://fedoraproject.org/wiki/Template:Infobox_user
* https://badges.fedoraproject.org/
* http://fedoramagazine.org/wp-content/uploads/2014/06/badges.png
* https://badges.fedoraproject.org/user/mattdm
--
Matthew Miller
<mattdm(a)fedoraproject.org>
Fedora Project Leader
9 years, 10 months