Inspirated

Screensavers like glmatrix have long been used by *nixers to woo people by showing them customizable animations as desktop wallpapers. Users of desktop environments such as Xfce have to however use xwininfo to determine and use the window IDs of their desktops (as the “-root” option stops working when the root window is overlayed by respective desktop managers e.g., Xfdesktop). For those who want to automate the startup process of XScreenSaver wallpapers in such environments, here’s a quick command you can use:

$ /usr/libexec/xscreensaver/glmatrix -window-id $(xwininfo -name "Desktop" | grep 'Window id' | sed 's/.*\(0x[0-9a-z]*\).*/\1/g')

Results:

(Click on the thumbnails for larger versions.)

Few *nix users are aware of existence of one /usr/share/dict/words on their machines. The original purpose of this file was to assist Unix programs in spell-checking. Now that every program that supports typo-prevention includes its own dictionaries, the words file no longer fares as something significant in the geek universe.

Nevertheless, the nifty gem can still serve as a fun place to find or coin new words based on lexicographical constraints. The omnipresent egrep command can be used to exploit the power of regular expressions against the English dictionary. Here’s how:

• Find all words containing 6 or more characters which don’t contain any vowel, dot or dash:

  -bash-$ egrep -i '^[^aeiou.-]{6,}$' /usr/share/dict/words

  -bash-$ egrep -i '^[^aeiou.-]{6,}$' /usr/share/dict/words

  bkbndr
  BSDHyg
  BSFMgt
  BSGMgt
  BSPhTh
  crwths
  crypts
  Cynthy
  Cynwyd
  cywydd
  flybys
  Flysch
  flysch
  ftncmd
  ghylls
  glycyl
  glycyls
  glyphs
  gypsyfy
  gypsyry
  Khlyst
  Khlysts
  Khlysty
  Kylynn
  kyschty
  lymphs
  lymphy
  Lynndyl
  MSGMgt
  mtscmd
  myrrhs
  myrrhy
  Myrvyn
  Myrwyn
  nymphly
  nymphs
  pgnttrp
  Phyllys
  Phylys
  phytyl
  psychs
  pyrryl
  rhythm
  rhythms
  Schwyz
  spryly
  SSTTSS
  stddmp
  strych
  styryl
  sylphs
  sylphy
  symphysy
  synchs
  synths
  syzygy
  thymyl
  trysts
  tsktsk
  tsktsks
  tyddyn
  vyrnwy
  why’ll
  Wrycht
  WWMCCS
  xylyls

• Find all words containing exactly 4 characters which can be spelled in pure Hexspeak, e.g., 0xDEADBEEF or 0xBABEFACE:

  -bash-$ egrep -i '^[abcdef]{4}$' /usr/share/dict/words

  -bash-$ egrep -i '^[abcdef]{4}$' /usr/share/dict/words

  AAAA
  AAEE
  abac
  Abad
  Abba
  abba
  Abbe
  abbe
  abed
  ACAA
  acad
  acca
  acce
  ACDA
  aced
  Adad
  adad
  Adda
  adda
  Adee
  AFCC
  affa
  Baba
  baba
  Babb
  Babe
  babe
  BAcc
  Badb
  bade
  BAEd
  baff
  bead
  Bebe
  Bede
  bede
  Beeb
  beef
  BFDC
  caba
  Cabe
  Caca
  caca
  cace
  CADD
  Cade
  cade
  CAFE
  cafe
  caff
  CDCF
  ceca
  Cece
  cede
  CFCA
  dabb
  Dace
  dace
  Dada
  dada
  Dade
  dade
  daff
  DBAC
  dead
  deaf
  debe
  decd
  deda
  dedd
  Dede
  deed
  Eada
  Eade
  EAFB
  Ebba
  ebcd
  ECAD
  ecad
  Ecca
  ecce
  EDAC
  Edda
  edda
  Edea
  edea
  Edee
  Faba
  Fabe
  FACD
  face
  fade
  faff
  FEAF
  Febe
  feeb
  feed
  feff

• Find all words which contain ‘H’, ‘T’, ‘M’ and ‘L’ in precisely that order:

  egrep -i '^h.*t.*m.*l$' /usr/share/dict/words

  egrep -i '^h.*t.*m.*l$' /usr/share/dict/words

  haemathermal
  haematothermal
  hemathermal
  hematothermal
  hepatoumbilical
  hephthemimeral
  heptametrical
  heteroecismal
  heteromeral
  heterothermal
  hexahydrothymol
  hippotomical
  histochemical
  histomorphological
  homeothermal
  homoiothermal
  homothermal
  hydrothermal
  hygrothermal
  hyperrhythmical
  hypersentimental
  hyperthermal
  hypertridimensional
  hypostomial
  hypothermal
  hysteromaniacal

• Find all words containing ‘s’, ‘e’ and ‘x’ but at least one different character between each of them:

  -bash-$ egrep -i '^.*s[^sex]+e[^sex]+x.*$' /usr/share/dict/words

  -bash-$ egrep -i '^.*s[^sex]+e[^sex]+x.*$' /usr/share/dict/words

  antispermotoxin
  asterixis
  Asteroxylaceae
  Asteroxylon
  Cristineaux
  Erysipelothrix
  erysipelothrix
  Herstmonceux
  Hurstmonceux
  inspectrix
  Issy-les-Molineux
  Lisieux
  mesoappendix
  obstetrix
  pressure-fixing
  proces-verbaux
  salenixon
  salpingemphraxis
  salteaux
  saucebox
  sauceboxes
  sceuophylax
  scleronyxis
  scleroticonyxis
  scleroxanthin
  she-fox
  side-box
  sidebox
  Sideroxylon
  single-tax
  skeptophylaxia
  skeptophylaxis
  slipper-foxed
  smokebox
  sneakbox
  sore-pressedsore-taxed
  sore-taxed
  spectatrix
  speculatrix
  spermatoxin
  spermotoxin
  sphacelotoxin
  sphenomaxillary
  spice-box
  splanchnemphraxis
  splenauxe
  splenotoxin
  state-taxed
  stenothorax
  sternomaxillary
  sternoxiphoid
  stone-axe
  Streptothrix
  subbureaux
  sulfadimethoxine
  superaxillary
  superfix
  superfixes
  superflux
  supergalaxies
  supergalaxy
  superluxurious
  superluxuriously
  superluxuriousness
  supermaxilla
  supermaxillary
  supermixture
  superoxalate
  superoxide
  superoxygenate
  superoxygenated
  superoxygenating
  superoxygenation
  supertax
  supertaxation
  supertaxes
  sweatbox
  sweatboxes
  swine-pox
  swinepox
  swinepoxes
  Thrsieux

Now you can name your start-up company “SupErfiX” and hope that it will someday be acquired by Microsoft.

The standard way of including source code listings in Beamer is to use the semiverbatim environment. Needless to say, it does not provide all the syntax highlighting and line-numbering love of the listings package. Combine the two and you have something pretty as well as extremely helpful in delivering presentations which have code:

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47

\documentclass{beamer}
 
\setbeamercovered{transparent}
 
\usepackage{pxfonts}
\usepackage{listings}
 
\begin{document}
 
\lstset{language=python,
        numbers=left,
        numberstyle=\tiny,
        showstringspaces=false,
        aboveskip=-40pt,
        frame=leftline
        }
 
\begin{frame}[fragile]
\frametitle{\texttt{parrot.py}}
\begin{semiverbatim}
\pause
\begin{lstlisting}
if __name__ == "__main":
 
\end{lstlisting}
 
\pause
\begin{lstlisting}[firstnumber=last]
    print "Oh yes, the, uh,
           the Norwegian Blue..."
    print "What's, uh...
           What's wrong with it? "
 
\end{lstlisting}
 
\pause
\begin{lstlisting}[firstnumber=last]
    print "I'll tell you what's wrong with it,
           my lad."
    print "'E's dead,
           that's what's wrong with it!"
 
\end{lstlisting}
\end{semiverbatim}
\end{frame}
 
\end{document}

While developing PyS60 apps is one of the most fun things you could do with your Nokia phone, debugging them isn’t as zippy as one would hope for in a Py development environment. To make up for that, PyS60 gives the developers an option for directly connecting to the interpreter through Bluetooth. Doesn’t sound very appealing? How about this: You connect your laptop with the cellphone, jump in at some place in the code while your app is executing and then use lappy’s big keyboard for exploiting different code and values in the interpreter. Sounds better?

To accomplish this on a Linux distro, you will need the following packages installed on your system:

After making sure that both are present on your system, install PyS60 on your phone if you haven’t already done so.

Now the fun part:

1. Switch on Bluetooth in the cellphone.

   

2. Launch bluetooth-properties and click on “Setup New Device”.

   

3. Select your cellphone.

   

4. You will be shown a pin.

   

5. Enter the pin when queried on the cellphone.

   

6. The phone should be successfully paired.

   

7. Authorize your Linux system to make automatic connections to the phone.

   

8. As root, run this shell script:

   [root@orthanc ~]# ./rfcomm-listen.sh

   [root@orthanc ~]# ./rfcomm-listen.sh

   Serial Port service registered
   Waiting for connection on channel 2

9. Launch PyS60 interpreter and select “Bluetooth Console” from the application menu.

   

10. Select your Linux machine.

    

    The command you ran in previous step should have new output:

    [root@orthanc ~]# ./rfcomm-listen.sh

    [root@orthanc ~]# ./rfcomm-listen.sh

    Serial Port service registered
    Waiting for connection on channel 2
    Connection from 00:17:4B:B6:35:31 to /dev/rfcomm0
    Press CTRL-C for hangup

    The cellphone screen should be showing something like this:

    

11. As root again, open a new terminal and run:

    [root@orthanc ~]# cu -l /dev/rfcomm0

    [root@orthanc ~]# cu -l /dev/rfcomm0

    Connected.

12. Hit Enter till prompt (>>>) appears, then type:
    

    >>> import appuifw
    >>> appuifw.query(u'Hello World', 'text')

13. Viola, you should have an input box on the mobile screen:

    

14. Enter any text and press the OK key. It should be show up in the terminal you were using to type in code:
    

    >>> import appuifw
    >>> appuifw.query(u'Hello World', 'text')
    u’Finally’

15. Exit the interpreter by typing CTRL+D on an empty line:
    

    >>> import appuifw
    >>> appuifw.query(u'Hello World', 'text')
    u’Finally’
    >>>
    Interactive interpreter finished.
    cu: Got hangup signal

    Disconnected.

Pat yourself on the back. Now, you can use your Bluetooth console to import your modules, execute some stuff and then jump in the middle to test some extra lines or values. In fact, I found it to be a pretty darned good way of learning about PyS60’s API. Res secundae!

I had never really laid my hands on LaTeX until I required it in one of the helper applications for my graduation project. Unfortunately, the requirement wasn’t as simple as producing some documents as I had to embed mathematical expressions on the fly in my PyGTK apps. Googling around for the solution, I found GtkMathView which accomplished something similar to this albeit using MathML. However, my luck ran out on me again as the widget lacked Python bindings. The other solution was to generate transparent PNGs on the fly and include them as GtkImages. This worked rather well, as the final code allowed easy modifications to the generated expressions.

Requirements for the code were:

• LaTeX
• dvipng
• latexmath2png

Final results:

And the simple code behind it:

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40

#!/usr/bin/env python
"""An example demonstrating usage of latexmath2png module for embedding math
equations in PyGTK
 
Author: Kamran Riaz Khan <krkhan@inspirated.com>
"""
 
import gtk
import os
import latexmath2png
 
pre = 'gtktex_'
eqs = [
	r'$\alpha_i > \beta_i$',
	r'$\sum_{i=0}^\infty x_i$',
	r'$\left(\frac{5 - \frac{1}{x}}{4}\right)$',
	r'$s(t) = \mathcal{A}\sin(2 \omega t)$',
	r'$\sum_{n=1}^\infty\frac{-e^{i\pi}}{2^n}$'
	]
latexmath2png.math2png(eqs, os.getcwd(), prefix = pre)
 
def window_destroy(widget):
	for i in range(0, len(eqs)):
		os.unlink(os.path.join(os.getcwd(), '%s%d.png' % (pre, i + 1)))
	gtk.main_quit()
 
window = gtk.Window()
window.set_border_width(10)
window.set_title('LaTeX Equations in GTK')
window.connect('destroy', window_destroy)
vbox = gtk.VBox(spacing = 10)
window.add(vbox)
 
images = [None] * len(eqs)
for i in range(len(eqs)):
	images[i] = gtk.image_new_from_file('%s%d.png' % (pre, i + 1))
	vbox.pack_start(images[i])
 
window.show_all()
gtk.main()

Recently, I needed to create some visualizations in a drawing area for one of my PyGTK apps. The PyGTK tutorial had an excellent writeup complete with a working example for immediate help. Unfortunately, my requirements were of a “grid” view on which other objects would be drawn. Naturally, I hoped for some variable on the GtkDrawingArea which I would set to true and have the grid lines drawn automatically. But since I couldn’t find any such magic setting in the API, I had to draw the lines manually.

Consequently, I could use either of the following approaches:

• Loop horizontally and vertically while draw_line()ing.
• Draw a 10×10 box as a tile and repeat it over the background.

The first approach would’ve worked faster but involved writing more code. The second was uglier — especially for larger images — but resulted in me having to add only a few lines to the example:

72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91

        xpm_data = ["10 10 2 1",
            " 	c #EEEEEE",
            ".	c #FFFFFF",
            "          ",
            " ........ ",
            " ........ ",
            " ........ ",
            " ........ ",
            " ........ ",
            " ........ ",
            " ........ ",
            " ........ ",
            "          "]
        tile_pixmap, tile_mask = gtk.gdk.pixmap_create_from_xpm_d(
            area.window, None, xpm_data)
        tile_gc = area.window.new_gc(fill = gtk.gdk.TILED,
            tile = tile_pixmap)
        width = int(self.hruler.get_range()[3])
        height = int(self.vruler.get_range()[3])
        area.window.draw_rectangle(tile_gc, True, 0, 0, width, height)

Turning this:

Into this:

Note that the 10×10 tiles’ borders match prettily with the markers on top and left rulers. There will definitely be some more efficient method for achieving this but for the time being tiling serves my needs perfectly.

Pidgin is a multi-protocol instant-messaging client which I use for all my MSN/Google Talk/IRC communication. Apart from offering adept support for all of these protocols (and plenty others as well), Pidgin also provides decent logging support. However, I’m not at all interested in having my daily MSN conversations recorded. On the other hand, I prefer having my previous IRC chats with me for reference. Right now, Pidgin does not support any options in the Preferences which would allow me to selectively turn on logging based on the protocols I’m using. If I turn it on for chat windows, it also starts producing logs for every MSN chat I participate in.

Most of the people wouldn’t/don’t see any issue with this behavior. For those who do want to keep a history of their conversations only for particular protocols, here’s a quick workaround for doing it:

• Check the appropriate options under the “Logging” tab in Preferences window.
• Have a few chats the likes of which you’d prefer being logged.
• Go to the directory containing Pidgin logs (default is ~/.purple/logs):

  [user@host ~]$ cd ~/.purple/logs/

  [user@host ~]$ cd ~/.purple/logs/

• See whether the protocols you want to be logged have a directory named for them:

  [user@host logs]$ ls

  [user@host logs]$ ls

  irc msn

• Remove the protocols you do not want to have logs for (in my case, MSN):

  [user@host logs]$ rm -rf msn

  [user@host logs]$ rm -rf msn

• Change directory permissions to stop new protocols from getting logged:

  [user@host logs]$ chmod 500 .

  [user@host logs]$ chmod 500 .

And you’re done. Now, whenever you start a conversation in a protocol which does not have a respective directory in ~/.purple/logs, you’ll see a “Logging failed” error message in the conversation window. For other (allowed) protocols, logging will work as expected. To turn off selective logging, reset directory permissions with:

[user@host ~]$ chmod 600 ~/.purple/logs/

It is also possible to apply the same workaround on contacts. For example … :

• [user@host ~]$ cd ~/.purple/logs/msn/ [user@host msn]$ ls

  [user@host ~]$ cd ~/.purple/logs/msn/ [user@host msn]$ ls

  bestbuddy@live.com ignorantmoron@live.com

• [user@host msn]$ rm -rf ignorantmoron@live.com

  [user@host msn]$ rm -rf ignorantmoron@live.com

• [user@host msn]$ chmod 500 .

  [user@host msn]$ chmod 500 .

… will disable logging for all MSN contacts including ignorantmoron@live.com, but will record everything communicated with bestbuddy@live.com. Like the previous example, you just have to reset directory permissions to re-enable nondiscriminatory logging:

[user@host ~]$ chmod 600 ~/.purple/logs/msn/

Every once a while, I need to take the backup of my Nokia N72 using PC Suite. Since the task had to be performed on Windows, I expected my virtualized machine to be able to do so. Unfortunately, Virtual Machine Manager does not provide any option in its interface which would allow me to use my USB devices in virtualized machines. Going through the documentation though, here’s the method through which I was able to solve my issue.

First of all, you should have the vendor and product ID’s of the USB device you want to use. Sounds alien? Use the command:

[user@host ~]$ lsusb

Which will show you something like:

Bus 001 Device 001: ID 1d6b:0002 Linux Foundation 2.0 root hub
Bus 005 Device 001: ID 1d6b:0001 Linux Foundation 1.1 root hub
Bus 004 Device 002: ID 0a12:0001 Cambridge Silicon Radio, Ltd Bluetooth Dongle (HCI mode)
Bus 004 Device 001: ID 1d6b:0001 Linux Foundation 1.1 root hub
Bus 003 Device 003: ID 0421:04c4 Nokia Mobile Phones
Bus 003 Device 002: ID 09da:000a A4 Tech Co., Ltd Port Mouse
Bus 003 Device 001: ID 1d6b:0001 Linux Foundation 1.1 root hub
Bus 002 Device 001: ID 1d6b:0001 Linux Foundation 1.1 root hub

The bold numbers in the line containing “Nokia Mobile Phones” are the vendor and product IDs respectively. Once you’ve noted them down for your required device (in my case: 0x421 and 0x4c4), list the virtual machines’ configuration files:

[user@host ~]$ sudo ls /etc/libvirt/qemu

networks windoze.xml

My virtual machine was named “windoze”, so windoze.xml is the file that I need to edit:

[user@host ~]$ sudo gedit /etc/libvirt/qemu/windoze.xml

In the editor, add the highlighted hostdev lines under the devices section (replacing the vendor and product IDs with the ones noted down from lsusb output):

<devices>
<emulator>/usr/bin/qemu-kvm</emulator>
<hostdev mode='subsystem' type='usb'>
<source>
<vendor id='0x0421' />
<product id='0x04c4' />
</source>
</hostdev>

Save and close the file. Restart the service:

[user@host ~]$ sudo service libvirtd restart

If everything went smoothly, the USB device should now be accessible from within the virtual machine:

(Click on the thumbnail for larger version.)

More often than not, Linux users will end up in a situation where they’d need to access their Windows neighbors on LAN. This can easily be accomplished using IP addresses, but using NetBIOS names is just too convenient to be overlooked.

The Xandros distro on Asus’ Eee PC does not resolve NetBIOS names by default. To make it perform that, you can do the following:

[user@host ~]$ sudo apt-get install samba samba-common smbclient winbind
[user@host ~]$ sudo kwrite /etc/nsswitch.conf

Once the editor opens, spot the line:

hosts: files dns

And append “wins” at its end, making it:

hosts: files dns wins

Save the file, exit the text-editor, back on command-line:

[user@host ~]$ sudo /etc/init.d/samba start
[user@host ~]$ sudo /etc/init.d/networking restart

Reconnect your network, and viola — easy peasy, Eee PC!

(Click on the thumbnail for larger version.)

I’m well aware of the fact that this thing has been posted on various blogs countless times already; but if you haven’t noticed, YouTube has changed its page structure so that only a text-search of player2.swf?video_id= in the source code won’t yield the prolonged video_id. Instead, you now have to search for /watch_fullscreen?video_id= to locate the holy grail.

For those who’re new to this method, here’s what you have to do:

• Open the YouTube video link for which you want to download the FLV.
• Right-click anywhere on the page and click on ‘View (Page) Source’.
• Press CTRL+F and search for the string: /watch_fullscreen?video_id=
• Copy everything after the question mark until the end of quotes, as shown in the image below:

  
  (Click to enlarge)

• Type this in your address bar: https://www.youtube.com/get_video?[video_id you've just copied]
• Hit Enter/Return, you’ll be offered a downloaded named get_video which is actually your desired FLV file. Save it somewhere locally and change its extension to .flv.

Here’s a list of players that you can use to play your downloaded FLV:

• MPlayer (personal favorite — cross-platform)
• VLC Media Player (cross-platform)
• ffdshow-tryouts codecs (Windows)
• FLV Player (Windows)

Bona fide :-) !