Speeding up app-office/libreoffice-bin generation

Since the last app-office/libreoffice-bin upgrade was long-delayed and overall slightly painful, we've modernized behind the scenes. Patrick Lauer a.k.a. bonsaikitten, who is amongst other things already running the commit notification bot irker on #gentoo-commits and the amazing AutoRepoman verifying quality and consistency of the portage tree, has provided me with generous access to two dedicated build chroots for amd64 and x86 libreoffice on a rather fast server. Together we're trying to automatize the upgrade process as much as possible. The app-office/libreoffice-bin-4.1.3.2 packages that I'm uploading right now and that will appear on the mirror servers and in the portage tree within the next 24h are the first result of the new setup. Test and enjoy!
Oh yeah, and since many people have asked me already, the awesome new splash screen was contributed by Denis M. a.k.a. Phr33d0m, see bug 456666. Thanks a lot too!

Gentoo on Facebook

Turns out Gentoo had a Facebook page with >10k "likes" already for quite some time. It was created in 2008 by a developer who later stopped using Facebook at some point, and noone really posted anything "official" there. Well, now's the time to step up public relations a bit... We (as in the people I spoke to on the #gentoo-pr Freenode channel) are not 100% sure so far what to do with it exactly, but there will certainly be occasional news and updates. Also, I'm feeding it with the Gentoo Planet RSS. If you are a Facebook user, head to the page and like us! :)

Lab::Measurement 3.20 released - including Lab::XPRESS

We have uploaded Lab::Measurement 3.20 to CPAN today. This release includes significant new features, in particular in the high-level measurement code and in the instrument readout logic.

The central new feature is Lab::XPRESS. This set of Perl modules builds upon the Instrument layer and provides a replacement and alternative to the Lab::Measurement class and its metadata handling. As detailed in several examples (1, 2, 3), nested loops of parameters, as e.g. gate voltages, combined with arbitrary measurements at each point can now be implemented in a highly flexible but simple way: you describe the logical structure of the measurement, without having to program the loops yourself. This means the actual measurement scripts become simpler and shorter, while remaining highly configurable. The underlying code can automatically take care of many details such as waiting times, timing of measurements, and pause or well-defined abort of the script on key events.

At the Instrument level, foundations have been laid for a centralized caching of instrument parameters. While this defaults off, it can be enabled in an instrument driver. The functionality is useful e.g. if reading out a parameter from the device that does not change on its own, unless explicitly set by a script, is a slow command. This also includes a generic support for asynchronous readout, i.e. first requesting a measurement value and thereby starting a read-out, then later in a separate command reading the result.

In addition, several instrument drivers have been added in varying states of completion, as e.g. the Cryogenic SMS magnet power supply, the Knick S252 voltage source, the Anritsu MG369xB signal generator, and the Keithley 2000 multimeter.

Finally, the TCPraw connection has been deprecated in favour of a new generic Socket implementation which provides a larger feature set. While TCPraw is still available, it internally only calls Socket and will be removed in a future version.

The release package is available from CPAN. Documentation as well as a web view of the Git code repository and a bug tracker are available from the package homepage https://www.labmeasurement.de/.

app-emulation/vmware-workstation-10 and friends

I finally got around to import the ebuilds for the new VMware Workstation and Player versions into the main tree. Thanks a lot to all the contributors on Bugzilla. Right now the ebuilds do not have any keywords yet since I have not tested the installation results at all myself, so if you want to try installing the software, remember you need a license for VMware Workstation, and 10 != 9 :], and then keyword it manually in /etc/portage/package.keywords as follows:

=app-emulation/vmware-modules-279* **
=app-emulation/vmware-workstation-10* **
=app-emulation/vmware-player-6* **

(Obviously this will only work on x86 and amd64, since the software only exists for these architectures.) Feedback of all sorts is welcome on the version bump bugs for VMware Workstation and the kernel modules and for VMware Player.

Not keeping up - DB ICE w-lan hotspots pointless

We used to be very much at the top of travel comfort, at least what network connectivity concerned. On many routes of the high speed trains within Germany, W-LAN hotspots were installed in the cars, and the network was (not free but) reliable and quite fast. (Which 3G mobile network connectivity usually is not once you're moving at 300kph.)
Now, in the age of the smartphone, unfortunately we learn firsthand what happens if supply does not follow demand... During each of my last train trips, the router was quickly out of IP adresses. Strong signal, wi-fi connection fine, no network layer. Which makes the whole thing extremely pointless. :(

pss(b) accepted: Negative frequency tuning of a carbon nanotube nano-electromechanical resonator under tension

One of the measurements in the MSc project of Sabine Kugler revealed a very particular carbon nanotube nano-electromechanical resonator. Usually the mechanical resonance frequency is tuned up to higher values by applying a gate voltage to the back gate: the larger charge on the nanotube and the larger forces between gate and nanotube cause mechanical tension within the nanotube, and exactly the same way as when tuning a guitar or piano string this translates to a higher resonance frequency. Curiously, here a very strong effect to the opposite can be observed; we can tune the resonance frequency to 75% of its zero-gate voltage value by applying a gate voltage! This is even more astonishing since we can conclude from the higher harmonics of our vibration mode that the nanotube is under tension.
The observed effect can be explained via so-called electrostatic softening of the vibration mode. Let us assume that the carbon nanotube is very close to the gate and vibrates towards and away from it. The capacitance between gate and nanotube varies within one oscillation cycle, and thereby the electrostatic force between these two obtains an additional position-dependent component. This can be seen as an electrodynamic contribution to the spring constant of the resonator; it is negative and thereby decreases the resonance frequency. We can estimate the size of this effect and obtain indeed consistent values for our sample geometry.

"Negative frequency tuning of a carbon nanotube nano-electromechanical resonator under tension"
P. L. Stiller, S. Kugler, D. R. Schmid, C. Strunk, and A. K. Hüttel
accepted for publication by physica status solidi (b), arXiv:1304.5092 (PDF)

Thank you

I'm going to keep this short: now that the election results for the Gentoo Council are out, thanks a lot for your vote of confidence. It'll be an honour and a pleasure to help making Gentoo the most awesome not-just-linux distribution ever. :)