Statistics
We looked inside some of the posts by micas-blog and here's what we found interesting.
Average Info
Notes Per Post
0
Likes Per Post
0
Reblog Per Post
0
Reply Per Post
0
Time Between Posts
7 months
Number of Posts By Type
Link
3
Text
9
Photo
1
Last Seen Tumblr Blogs
Fun Fact
12.7% of mobile users access Tumblr.
Link
If you like to configure your Magic Mouse to single finger tap as the trackpad does, this is the tool you need.
If you search for a tool with even more features like window snapping or keyboard shortcut configuration or controlling your Mac's mouse from remote by iphone/ipad try https://www.boastr.net/
0 notes
Text
Using HG with GERRIT
If you like HG and are forced to use GERRIT, because it's the code review tool that must be used, so there is help to workaround the GIT-hell: Stay with HG and use the HG-GIT extension.
Install the extension once like in http://mercurial.selenic.com/wiki/HgGitand enable option intree, to get a ".git"directory on the same level as the ".hg" directory.
In $HOME/.hgrc or mercurial.ini file add the following line:
[git]
intree=1
And here is, how it works:
develop in your workarea and repo, create commits
add a bookmarks for your change - HGs bookmarks work much the same like GIT branches add a master bookmark - because GIT-people like it: hg bookmark master -r default add a feature bookmark: hg book feature1
as you commit patches, the bookmarks moves with your latest commit
export HG-changes into a temporary GIT-repo: hg gexport
push changes to gerrit git push ssh://user@gitserverhost:29418/your-repo feature1:refs/for/master
if you get an error about a missing "Change-Id" please add it to the commit message, using e.g. commit with amend option of the MQ-patch mechanism.
0 notes
Text
Why HG is better than GIT
Because HG (a.k.a. Mercurial Version Control System) does things simply clever. There is usually one easy way to do it and it just works.
Example? Create a bare repository (one with no workarea) and sync regularly changes from all branches.
With HG
To setup: hg clone -U <dir1> <dir2>
and regularly: hg pull
That's it. This gets all changes, new bookmarks (equivalent to GITs branches).
With GIT
... to complex. :-)
To setup:git clone --mirror <dir1> <dir2>
So you get a mirror of the repository with "remotes" and "remote refs" setup, but there is no easy way to get all new git branches from <dir1> without knowing their names.
There is also "git fetch --all", but this doesn't fetch (get) all changes, it just ask every remote location, that was setup for this repository.
0 notes
Link
0 notes
Text
Formel 1 ohne Teamorder
Um die aktuelle Diskussion um eine Teamorder zu l枚sen habe ich folgenden Vorschlag.
Neue Regel bei Punktevergabe f眉r alle Motorsportserien mit Fahrer- und Teamwertung wie z.B. Die Formel 1.
Die Punkte werden wie bisher nach dem Platz beim Zieleinlauf vergeben. Falls die Fahrer eines Teams auf unmittelbar aufeinander folgenden Pl盲tzen ins Ziel kommen, erh盲lt jeder Fahrer den Durchschnitt der beiden Platzierungen. Eine Ausnahme gibt es bei der Platzierung von Teamkollegen auf 10 und 11, wo beide Fahrer je 1 Punkt erhalten und das Team damit 2 Punkte bekommt. Beispiel aus der Formel 1 bei Einlauf auf den Pl盲tzen 1 (25 Pkt.) und 2 (18 Pkt.) erhalten beide Fahrer jeweils 21.5 Punkte (43/2).
Folgen: 1. Team-Kollegen Positionstausch ist nicht notwendig, wenn beide hintereinander fahren. Es ist nicht wichtig, sich gegen den Teamkollegen durchzusetzen, wenn keine weiteren Positionen erk盲mpft werden k枚nnen. 2. Gleich gute Fahrer in einem Team k枚nnen die gleiche Punktezahl in einem Rennen bekommen. Fahrer, die sich mehr als einen Platz vom Teamkollegen absetzen, erhalten mehr Punkte und werden f眉r eine bessere Leistung auf gleichem Material belohnt. 3. Will ein Team einem Fahrer mehr Punkte geben als dem anderen, weil z.B. einer in der Meisterschaft vorn liegt, muss es den 2. Fahrer einen Platz verschenken lassen, damit ein anderes Team zwischen den beiden Teamkollegen ist. Dabei muss das Team Teampunkte verschenken und wird dadurch f眉r diese Fahrerbevorzugung bestraft.
0 notes
Text
wxPython for Python 3 on Mac
Download snapshot build of wxPython as described on PythonLibrary Blog [1] from wxPython Download [2]
Start the python3.2 interpreter and run the following commands to find out your site-package directory:
import sys print(sys.path) for p in sys.path: if "site-packages" in p: print p
Look for the path with 'site-packages' and move the wx directory to that folder. E.g:
sudo cp -r wx /opt/local/Library/Frameworks/Python.framework/Versions/3.2/lib/python3.2/site-packages/
Try importing wx into your python shell:import wx
[1]http://www.blog.pythonlibrary.org/2012/05/24/wxpython-for-python-3-is-almost-here/ [2]http://wxpython.org/Phoenix/snapshot-builds/
0 notes
Text
Developer tools on Mac OSX Lion
After installing Xcode on Mac OSX Lion the developer tools like make, gcc and so on could not be found in /Developer/usr/bin as usual.
To find this tools again in your PATH you could follow these steps:
Install Xcode (from App Store)
Add the file "/etc/paths.d/20-Xcode" as admin user and
add the line "/Applications/Xcode.app/Contents/Developer/usr/bin/" and save the file.
Open a new terminal and you will find make, gcc and friends.
0 notes
Text
Der erste Mac
F眉r den Anfang mit dem neuen Mac sind bei Apple einige Videos zu finden.
Vom PC zum Mac
Ein Mac im 脺berblick聽
und weitere unter聽http://support.apple.com/de_DE/videos
0 notes
Photo
What's an application? Something nice looking put on clothings even if it runs on a multimedia player.
0 notes
Text
GNU Makefile dependency on clean target
If your Makefile includes an "all" and a "clean" target, it surely runs fine on a rebuild with:
make clean all
But what happens if it should speed up build with parallelisation and is started with option -j?
make -j 8 clean all
Now it could happen, that the building step (all target) is already executed, while "clean" is still running. E.g. all is creating files into build directory, while clean is removing this directory. You get build errors running in parallel mode, but this is not the fault of make. There is no dependency from "all" to "clean".
# Makefile clean: 聽聽 @echo clean ... 聽聽 @sleep 5 聽聽 @echo clean done
all: clean 聽聽 @echo "all ..." 聽聽 @sleep 3 聽聽 @echo "all done"
Fine. This is working very well, if we run "make clean all". The target "all" is executed after "clean" is finished. But if you run only "make all" - target "clean" is executed too, because "all" depends on it. That's not what we wanted.
But there is a little trick and here it comes. GNU make has a variable called MAKECMDGOALS which includes all targets specified on commandline. This variable can be searched for "clean" and only in this case, the dependency from "all" to "clean" is made. Let's look at it:
ifeq ($(findstring clean ,$(MAKECMDGOALS) ), clean ) 聽聽聽 CLEAN_DEP:=clean endif
all: $(CLEAN_DEP) 聽聽 @echo "do building"
clean: 聽聽 @echo "do cleaning"
If this tip was usefull, you could leave a comment on this site.
0 notes
Text
Versionsmanagement mit Mercurial
Einfacher Einstieg: http://hginit.com
Branching in Mercurial: stevelosh blog, also there, best-practices
workflow with bookmarks
0 notes
Text
Openshot Video-Editor
Und wieder eine Videoschnittsoftware f眉r Linux. Einen Test ist sie sicherlich wert. http://www.openshotvideo.com
0 notes