wulfovitch.net

iTunes Connect and Umlauts

2010-05-04T17:08:00+02:00

Apple is a american centric company, otherwise it is not explainable that you can get a lot problems if your name does not consist solely of ASCII letters. My last name - König - contains a so called umlaut: an “ö”. Everytime apple posts a new license agreement to the developers, the developer has to accept this new license agreement if he wants to continue to develope applications for the iPhone. This agreement affects also the iTunes Connect portal. The iTunes Connect portal is the place where the developer can manage his applications which he wants to put on the app store.

And here the problem begins: if the developer accepts a new agreement in the developer portal, the status of this new agreement “should” also be updated in the iTunes Connect portal. It doesn’t, if your name does contains umlauts like my last name does. The developer portal states that I have accepted the new license agreement, the iTunes Connect portal doesn’t. The consequence is, everytime apple posts a new license agreement, I have to contact the iTunes Connect support and have to wait until they set the correct bit in their database manually. This happened to me already five or six times and it is starting to get annoying as you can think.

Now I have noticed that my application in the app store has been dropped silently, because iTunes Connect thinks, I haven’t accepted the new license agreement. I don’t even know for how long the application isn’t available in the app store anymore, because apple gave me no indication that it isn’t. It is really embarrassing for a huge company like apple, that they can’t even handle encodings correctly in their developer portal and moreover they should sent out status messages to the developer if the availability of an app in the app store changes.

Update - 10 May 2010

It took Apple six days to resolve the problem. Now my application is available again in the app store. At least this has happened automatically.

Jekylless

2010-02-16T21:33:00+01:00

Tate Johnson recently created his own fork of Jekyll named tatey-jekyll. He added the support for LESS CSS and growl notifications to it. I liked the idea and merged the latest changes from Jekyll into his fork.

While I was at it, I also added the support for specifying the concrete time for a blog post in the yaml from matter. Whenever do you publish blog posts at twelve o’clock at night?

Tate liked the changes I made and we’ve agreed that we will continue to develope this fork under the name Jekylless. The name Jekylless was chosen to show the projects’ support for LESS CSS.

We both will integrate a few more features into this fork, but we’ve decided to stay as close as possible to the codebase of the original Jekyll, so that switching to our fork should be easy for every existing jekyll user.

For more information about Jekylless, checkout the readme of the project on github or install the Jekylless gem directly.

Dropzone SCP Script with Gallery Creation

2010-01-21T21:34:00+01:00

Dropzone is a great tool for all your drag & drop needs on a OS X system. Also cool about dropzone is, that you can extend it to your personal needs. I’ve always wanted a tiny little helper where I can drop my files onto. A gallery should automatically be created from these files and this gallery should be uploaded to my webspace afterwards. This kind of script would especially be useful if I wanted to show a friend some pictures rather quickly. I’ve never found a really good working solution for this on the web. At the end, I’ve created this dropzone script which does what I needed:

This script takes care of a few things:

If there is more than one file dropped onto the script, a gallery is created automatically.
Big pictures are being resized to a reasonable size.
Files which are not pictures are only getting linked to.
You can enter a name for the gallery, if you are pressing the ‘Control’ key when dropping the files.
After the upload, the url of the file or the gallery is pasted to the clipboard.

You need to have ImageMagick installed, if you want to get the resizing to work. The easiest way to install it, is via MacPorts and the following terminal command:

sudo port install ImageMagick

If you don’t have ImageMagick installed, the pictures will be uploaded as is without resizing.

You can obtain the script at github or download it here.

Update

The script resides now at the official dropzone user scripts repository, too!

LaTeX Compile Script

2009-12-22T17:19:00+01:00

I am dealing with a lot of LaTeX documents lately. The need to compile every tex document at least two times is really annoying. And if you also have to use bibtex, then it gets real cumbersome to compile all the appropriate files properly. Though I have written this small ruby script, which takes care of all the files and compiles them in the correct order:

#!/usr/bin/env ruby
# default values
compileCount = 3
open_latex_document = false

def outputErrorMessage
  puts "Too many parameters..."
  puts "Usage: 'rubylatex.rb anylatexfile.tex [compilecountnumber]'"
  puts "where [compilecountnumber] is a number greater than 0"
end

# check arguments
if ARGV.count == 1
  latexfile = ARGV[0]
elsif ARGV.count == 2 || ARGV.count == 3
  latexfile = ARGV[0]
  compileCount = ARGV[1].to_i
  
  # check if compileCount has a reasonable value
  if compileCount < 1 || compileCount > 5
    outputErrorMessage
    exit
  end
  
  # open the latex document if a third parameter has been detected
  if ARGV.count == 3
    open_latex_document = true
  end
else
  outputErrorMessage
  exit
end

# filename without file type extension
filebasename = File.basename("#{latexfile}", '.tex')

# compile latex file at least one time 
sout = system("pdflatex #{latexfile}")

# create bibtex files
if sout
  sout = system("bibtex "+filebasename)
end

# compile latex files a few times more
if sout && compileCount > 1
  (2..compileCount).each do |i|
    if sout
      sout = system("pdflatex #{latexfile}")
    end
  end
end

# cleanup
puts "#### cleaning up! ####"
system("rm -rf *.log")
system("rm -rf *.aux")
system("rm -rf *.aux.bak")
system("rm -rf *.toc")
system("rm -rf *.out")
system("rm -rf *.idx")
system("rm -rf *.blg")
system("rm -rf *.bbl")
system("rm -rf *.nlo")
system("rm -rf *.lot")
system("rm -rf *.lof")
system("rm -rf *.tcp")
system("rm -rf *.tps")
system("rm -rf *.brs")
system("rm -rf *.brf")

# pdf successfully generated
if sout
  puts "#### done! #{filebasename}.pdf is ready! ####"
  if open_latex_document
    puts "#### opening #{filebasename}.pdf now! ####"
    system("open #{filebasename}.pdf")
  end
else
  # failed pdf generation
  system("rm -rf #{filebasename}.pdf")
  puts "#### failed! no pdf generated!!! ####"
end

The script is a little bit bloated, but it gets the job done and it even takes care of deleting all the temporary files which are created during the compiling process.

Change Terminal Colors on Remote Connections

2009-12-11T10:57:00+01:00

I often use ssh in order to work on remote servers. But often it is hard to tell at first sight, if a open terminal window displays the contents of remote connection or if it is a local one. Therefore it would be nice, if the terminal would automatically switch colors, when a remote connection is detected. On Stack Overflow I found a solution for Terminal.app on OS X, which I adapted and improved a little bit:

#!/bin/sh
HOSTNAME=`echo $@ | sed s/.*@//`

# define background colors
BLACK_BG="{0, 0, 0, 50000}"
GREY_BG="{10000, 10000, 10000, 50000}"
RED_BG="{10000, 0, 0, 50000}"

# define font colors
GREEN="{0, 65535, 0}"
LIGHT_GREY="{55000, 55000, 55000}"


set_bg () {
  osascript -e
  "tell application \"Terminal\" to set background color of window 1 to $1
   tell application \"Terminal\" to set normal text color of window 1 to $2"
}

on_exit () {
  # set colors back to normal when quitting the remote connection
  set_bg "$BLACK_BG" "$GREEN"
}
trap on_exit EXIT

case $HOSTNAME in
  # you can set your production server in your .bash_profile file
  # like so: export production1="example.com"
  $production1|$production2|$production3) set_bg "$RED_BG" "$LIGHT_GREY";;
  *) set_bg "$GREY_BG" "$LIGHT_GREY" ;;
esac

/usr/bin/ssh "$@"

I’ve put this script into my ~/bin/ folder, which is the first place where my scripts are looked up. So don’t forget to add this folder to your $PATH. Everytime a remote connection with ssh is started, this script is first called, changes the colors of the terminal and issues the real ssh command.

About URL-Shorteners

2009-12-06T21:33:00+01:00

URL-Shorteners are getting popular these days, due twitter’s 140 character limitation. But what happens behind the scenes, if you are opening a url, created by an url-shortener? The shortener takes the request from the browser, looks up the given value of the short url in the database and redirects the browser to the original site. And here is the point, where the trouble comes in:

A short-url is not human-readable and therefore all semantic information is lost. You have no idea, to which url the short-url is leading you to and maybe you are being refered to a site you don’t wanted to visit at all.
If the url shortener service is shutdown or not reachable for some reason, you have no idea to what url the short-url was pointing to.
In some cases, some url-shortening services redirected old short-urls to malware sites or something like that. If you or somebody is looking up some old tweet of you, the short-url has become useless although the original maybe still available.

Short-urls are adding a layer of indirection to the internet, which is not needed in my opinion.

What could Twitter do about short-urls?

People are using url-shorteners, because they are forced to limit themselves to 140 characters. Twitter could implement a different counting mechanism:

A url is counted as 20 characters no matter how long the url is in reality
Limit the maximum number of urls in a tweet to 3.

By using this behaviour, no one has a real need for short-urls anymore and you will always know, how the real url looks. It should be fairly easy to implement this for twitter, but i think they are not doing this, because people cheat to get more characters in a single by simple typing something like: “http://and-i-also-wanted-to-say-this-and-that.com”. In my opinion, this would be acceptable tradeoff if we could get rid of url-shorteners! You could just unfollow the people who are abusing the now legal long urls for something else and you are done.

My current solution

Because I don’t think twitter will solve the problem with url-shorteners in the near future, I have set up my own with the help of Shaun Inman’s lessn. This solution won’t help to get rid of short-urls, but I am at least no longer dependent of a third party url shortener. I am in control of the short-urls I have posted on twitter and so I will never ever lose a link again, because a short url service abused my trust in their service.

My New Blog

2009-11-29T23:12:00+01:00

So, this is my new blog. In fact, my only blog before was a posterous account which I used as a blog¹. But posterous wasn’t the blog engine I was looking for, so I finally decided to host the blog by myself. But what blog engine should I use? I could use WordPress like everyone else does. There is nothing wrong with WordPress, but it doesn’t fit my personal requirements. These requirements are:

I don’t want to use a blog system where I have to take care of updating the blog engine itself constantly. In most cases, this is the case, when a new security hole was discovered.
I don’t want to fiddle around with plugins or something like that to get something running.
I want to write blog posts in markdown.
I want a blog engine, which doesn’t need much resources.
I want to have syntax highlighting for code snippets.
I want to be able to backup my blog posts with no special effort.

Two months ago, I found out that I wasn’t alone with these requirements. I stumbled upon jekyll, a static site generator written in ruby. “A static site generator? That sounds so like the 90s!” you may think. But in fact, a static site generator is the perfect system for a blog engine. Think about it, what makes a blog a blog? A blog consists of one or more blog entries, which are all consis of a title, a date and of the text of the blog entry itself. Nothing more. If you write only a few blog posts in a month, the content of the blog itself doesn’t change very often. A perfect use case for a static site generator! Write the new blog post, generate the site, and upload this site with the tool you want. You don’t even need a server with special capabilities! The server only needs to serve static html files. So you don’t even have to worry about a special caching system, because caching is already given with static files!

Advantages of jekyll

With a generated static site, I don’t have to worry about keeping the blog engine itself up to date. I don’t have to fiddle around with plugins. With jekyll I can write blogposts in markdown. Syntax highlighting of code is also supported. Every single blog post is saved as a single text file, so I can easily back them up with my version control system of choice, which is git by the way. I don’t have to worry about backing up the database. Because no database is needed at all!

Final thoughts

I have now written more than I thought. Perhaps you want to comment to this blog post. This is the only thing, you can’t support with a static site generator like jekyll. But what you can do, is to host the comments externally. For this purpose I use the disqus commenting system, which I think is great, because they handle all the anti-spamming stuff for me. One thing more, I don’t have to worry about. The less I have to worry about, the more I can be lazy. Which is a good thing, because all good programmers are programmers because they are lazy! ;)

If you find it, you can keep it! :)
↩