PyDev, Python and system default Unicode encoding problem

Python 2 has a thing called “default encoding” to automagically encode Unicode strings when they are presented as byte strings. This is evil and has been discussed various times before.

What could be even more evil? Something in your development environment messes this setting set for you, without telling you that. This way you never encounter Unicode problems on your development computer and when you roll out your seemingly working code to production, the world goes haywire.

Evil. Evil. Evil. Thousands of curses and overworking hours to fix the problems.

I encountered this problem. And this is the code I used to track the problem down in site.py:

# Trap the bastard messing with the default encoding
# using a monkey patch
old_set_default_encoding = sys.setdefaultencoding

def aargh(x):
    import pdb ; pdb.set_trace()

sys.setdefaultencoding = aargh

--Return--
> /home/moo/py24/lib/python2.4/site.py(485)aargh()->None
-> import pdb ; pdb.set_trace()
(Pdb) bt
/home/moo/py24/lib/python2.4/site.py(613)?()
-> main()
/home/moo/py24/lib/python2.4/site.py(604)main()
-> execsitecustomize()
/home/moo/py24/lib/python2.4/site.py(514)execsitecustomize()
-> import sitecustomize
/home/moo/Desktop/Aptana Studio 2.0/plugins/org.python.pydev_1.5.3.1260479439/PySrc/pydev_sitecustomize/sitecustomize.py(99)?()
-> sys.setdefaultencoding(encoding) #@UndefinedVariable (it's deleted after the site.py is executed -- so, it's undefined for code-analysis)
> /home/moo/py24/lib/python2.4/site.py(485)aargh()->None
-> import pdb ; pdb.set_trace()

--Return--> /home/moo/py24/lib/python2.4/site.py(485)aargh()->None-> import pdb ; pdb.set_trace()(Pdb) bt  /home/moo/py24/lib/python2.4/site.py(613)?()-> main()  /home/moo/py24/lib/python2.4/site.py(604)main()-> execsitecustomize()  /home/moo/py24/lib/python2.4/site.py(514)execsitecustomize()-> import sitecustomize  /home/moo/Desktop/Aptana Studio 2.0/plugins/org.python.pydev_1.5.3.1260479439/PySrc/pydev_sitecustomize/sitecustomize.py(99)?()-> sys.setdefaultencoding(encoding) #@UndefinedVariable (it's deleted after the site.py is executed -- so, it's undefined for code-analysis)> /home/moo/py24/lib/python2.4/site.py(485)aargh()->None-> import pdb ; pdb.set_trace()

Looks like the culprint was PyDev (Eclipse Python plug-in).  The interfering source code is here. Looks like the reason was to co-operate with Eclipse console. However it has been done incorrectly. Instead of setting the console encoding, the encoding is set to whole Python run-time environment, messing up the target run-time where the development is being done.

There is a possible fix for this problem. In Eclipse Run… dialog settings you can choose Console Encoding on Common tab. There is a possible value US-ASCII. I am not sure what Python 2 thinks “US-ASCII” encoding name, since the default is “ascii”.

Creating a drag’n'drop basket with jQueryUI

I have created an example how to create a “basked” with jQuery and jQueryUI with the following features

• The user can be pick items from the predefined set
• Items are dragged and dropped to the basket
• The basket value reflects a hidden input

This kind of user interface pattern is suitable for

• Shopping carts
• Travel planners

The example in fact bears the name “travel planner” but it is not tied to travel anyhow.

Note: this is just a screenshot - please see live example

The example code uses

• google.code() content delivery network to load jQuery and jQueryUI
• jQueryUI draggable and droppable features
• Well planned visual cues for the drag and drop operations: cursor changes, CSS hover classes
• <form> which <input> value is updated according to the basket content – all selected item ids form a comma separated list

The example code is well-documented with links to the further documentation.

• Live example
• View source code

Mobile browser wars: Nokia microB vs. Firefox Fennec

Begun, this mobile browser war, has. When mobile internet is growing 8x faster than desktop internet everyone wants to have a share of it. In the core of this fight is the mobile browser – the doorway to the mobile internet.

Usually phone comes with a browser from the phone manufacturer: Safari ships with iPhone, Android ships with WebKit based browser and Maemo comes with Nokia microB. Besides the default browser, open platforms have seen third party browsers created for them: Opera Mini has 30 million users and several browsers have been created for Symbian platforms. (Note that iPhone is not really open platform regarding this as Apple developer terms specifically forbid creating alternative  browser engines for their Safari – all iPhone “browsers” are just the same Safari with new toppings).

Now Mozilla foundation is releasing Firefox Fennec (RC1 version is available  for Nokia N900), touted as the most innovative mobile browser this far. New user interface ideas, desktop syncronization and vibrant add-on community are something yet to be seen for mobile browsers. Mozilla did an amazing thing with Firefox when it actually managed to push Internet experience forward and compete against Microsoft’s bundled Internet Explorer with sheer quality. Can Mozilla repeat the same thing it did for desktop browsing for mobile browsing too?

Is Fennec good? I installed the release candidate and conducted some tests by visiting on popular sites. It is especially fruitful to compare Fennec against Nokia’s own microB browser as they both are based on the same Gecko rendering engine beneath the hood.

The differences of the browsers are, actually surprisingly, not limited to branding and user interface shell. Fennec is portable browser – Mozilla hopes to run Fennec on other mobile platforms beside Maemo in the future. Fennec user interface is based on Mozilla’s XUL library and you can actually run Fennec on your desktop computer too. Nokia’s interest, on the other hand, is have an optimized browser for their own mobile phones: microB user interface is using native Maemo user interface components.

Below are some aspects of the browsers compared against each other.

Start up time

• microB: instant
• Fennec: About ten seconds (warm start-up is little bit faster, but it is still slooooow….)

This is a pain for Fennec. Loading all that XUL Javascript  needed to run Fennec is just too much. You really don’t want to start Fennec for a quick browsing session, unless you have the patience of a cow. I am not sure whether N900 keeps microB loaded on the background all the time or what’s causing the difference.

User interface

This is really where Fennec shines. Nokia enjoys some reputation of being a boring engineer house with little innovation left to stir. After learning the trick of left and right sweep, which is cleverly demostrated on the start page, Fennec user interface instantly feels intuitive. microB, on the other hand, uses somehow clumsy “bottom right corner full-screen button” to access buttons and left-right sweep is not very well thought. For example, switching a tab/browser window takes three “clicks” on microB (show menu – switch application – choose next browser window) when Fennec does it with one sweep and click. Also, backward navigation is much more intuitive on Fennec and takes too many gestures on microB.

Both browsers have search integrated to the navigation bar. Fennec start screen is more clever, showing the history and shortcuts, while microB shows only the bookmarks.  Fennec navigation bar also is a combination of title and navigation bar, saving the precious screen estate on small physical form factor. Fennec zooms to text fields automatically when you start to input text into them and also have soft “tab keys” to navigate to next and previous input field.

Page reading and speed

On sites with above average layout complexity, Fennec is unbearable slow compared to microB, up to the point the browser is next to unusable in its current incarnation. As they both use the same rendering engine, I have hard time to understand how microB manages even the heaviest dynamic pages (Facebook profile page) when Fennec becomes unusable even on a moderate complex page (slashdot.org).

The thing with Fennec is that for some of the the time it does not register your interaction and does not have any indicator showing if it is responding – it has grinded to halt, little bit like desktop computer when swapping.  And even when Fennec is responding the scrolling of the page refreshment is sluggish compared to microB. This makes the page reading experience unusable. A normal user won’t stand 1-3 second frequent responsivity pauses or page movement which cannot be controlled.

microB must do the rendering somehow different  - is it hardware acceleration on font rendering, smarter management of images or some other trick?. However, until Fennec reaches the smoothiness of microB, there is no way I would switch to Fennec over microB.

(Note: You can press CTRL-Backspace from N900 keyboard to force application switch if you cannot exit from halted Fennec otherwise)

Mobile browsing

Though N900 has 800 pixel wide screen, it is still a mobile phone. Small physical size, low bandwidth with high latency and  limited CPU power might make you to pick a mobile internet version of the site when it is available. However, since the screen has exceptional high Dots-Per-Inch value, this poses a problem for rendering sites with the default font sizes.

Fennec does not seem to have a shortcut for setting a large text size. This is something one would hope to see on such high DPI device as the most of the time default web site fonts are too small to be usable. Also, Fennec does not use the shoulder plus and minus volume buttons for zooming – microB does it and it is very natural place for this function.

Fennec seems to have some difficulties with mobile site rendering: for example touch.facebook.com and yle.mobi are not scaled to full width. Instead a narrow colum of 1/3 screen width is displayed.

Bugginess

microB is very solid piece of software. It crashes more rarely than Safari on iPhone (might this be because of more memory – low memory conditions seem to be a normal crashing condition for Safari?).  Fennec is still in its first version and have some issues.

(Note: I managed to get Fennec to zombie state – I had to go to terminal and type killall fennec command to make the browser become launchable again).

Sites tested

Slashdot.org

Geek discussion site

microB: no problems

Fennec: slow, frequent pauses, not smooth scrolling

slashdot.org/palm

Very simple mobile version of the above.

microB:  Font too small

Fennec: Scales correctly

Facebook.com

High profile social networking site

microB: Sometimes little slow, but seems to work perfectly

Fennec: Unusable slow

touch.facebook.com

microB: Perfect (at least when scaling font up a little)

Fennec: Does not scale correctly (default scale uses only 1/3 of screen width, double click zooming scales too much)

yle.fi

Finnish national broadcasting company site

microB: Ok. Readable and usable with text size large.

Fennec: Ok. The default view is navigable, but not readable. You need to double-click zoom to read the text (Fennec doesn’t seem to have text size large option)-

yle.mobi

The mobile version of above.

microB: Perfect with text size large, ok otherwise (need to double click to zoom and then click to choose a link to follow).

Fennec: Ok – font size too small

GMail HTML version

The default Javascript version of GMail is too heavy for both the browsers. GMail still provides “Basic HTML” view as the fallback for devices with less CPU power and network bandwidth.

microB: Ok – you can do some basic emailing

Fennec:  Ok. Does not seem to be affected by as much of “slugginess” as other sites are (might the slugginess be a Javascript issue?)

Youtube.com

The web version of flash based video sharing site.

microB: Plays Flash movies ok – smooth scrolling even whilst a Flash movie is playing

Fennec: Frequent grinds to halt, sluggish, unusable. Manages to open Flash video, though.

m.google.com/youtube

The mobile version of above.

microB: Youtube claims the browser is unsuppoted

Fennec: Cannot enter the site – shows only the page of Youtube Mobile instructions

twitter.com  (web site)

microB: Perfect

Fennec: Ok. Sluggish when opening new pages, but still usable. Fennec start view ships with Twitter button, so one might assume this site is well tested for Fennec.

m.twitter.com

The mobile site of above.

microB: Ok – the default font size too small, but when settings text size large works well

Fennec: Ok – the default font size too small. Double click zoom does not work well on the twit feed, making reading difficult.

plone.org

A community site with relatively simple layout.

microB: Ok – minor rendering errors

Fennec: Ok – minor rendering errors

iltalehti.fi

Finnish tabloid web site with lots of images.

microB:  Ok

Fennec: Grinds to halt, unusable slow

Summary

Though having nice promise of innovation, the advise for Fennec development team would be “back to the basics”. The slugginess and response times of Fennec are such an issue that one would not yet consider it as an real alternative for Nokia’s default microB browser.

With Fennec’s user interface and microB’s speed one could have a near perfect mobile browser. Depending what kind of future co-operation Nokia and Mozilla foundation will have, we might live to see it.

Profiling PostgreSQL database

This blog post will have some short notes about monitoring and profiling PostgreSQL databases.

pgtop

pgtop provides UNIX top command like user interface for PostgreSQL. pgtop command is available as Perl CPAN module.

How to install Perl CPAN modules as non-root user on Ubuntu (note: when it prompts to run sudo, answer no).

To install pgtop install following CPAN modules first: Term::ANSIColor, Term::ReadKey, DBD::Pg

perl -MCPAN -Mlocal::lib -e 'CPAN::install(DBD::Pg)'
perl -MCPAN -Mlocal::lib -e 'CPAN::install(Term::ReadKey)'
perl -MCPAN -Mlocal::lib -e 'CPAN::install(Term::ANSIColor)'

pgtop install instructions

pgtop manul

Running pgtop:

perl pgtop -d databasename -u yourdbuser -p yourdbuserpassword

pgfouine

pgfouine is a log analyzer for PostgreSQL.

Nokia N900, sports tracking and geotagging

This blog contains some tips how to use your Nokia N900 smart phone as a “augmented reality” sports device.

Sports tracking

Sport tracking is about collecting your sports activity data using GPS and other equipment. After running/cycling/skiing/whatever you see where you have been, how much time it took and how fast you are. In same cases you are able to calculate burnt calories and estimated heart rate.

N900 has at least one sports tracking application out there, eCoach.  eCoach is also suitable for professionals as it has heart rate monitor integration.

eCoach allows you record and  store sport activities. During the activity it uses Open Street Map based map viewer to show your current location. At least Helsinki area has very detailed maps available there, showing even the smallest trails, so you can safely venture to unknown neighbourhoods.

eCoach exports its tracks as GPX gps data file format. eCoach does not have any service integration yet, but you can upload this file to Nokia Sports Tracker and Map My Tracks. The recommend the latter as it has better social media integration and seems to be under active development. On the otherhand I have been using Nokia Sportstracker since 2007 and it has not really development during the whole this time and seems to lack will to go forward. Also Nokia has disabled track profile for imported GPX files which gives a message “we really don’t care about this service”.

There is also a service called mapmyrun.com with various domain names like “maymysomething.com”. Steer away from this service as I tested it and it didn’t live up to my expections (too much advertising, horrible user interface).

Some sport tracks I have made

• Run in Map My Tracks
• Another run in Map My Tracks
• The same run in Nokia Sportstracker
• All my runs in Nokia Sporstracker

Geotagging

Geotagging is about having GPS  coordinates on your photos. This way photos can be put on the map autotically in photo sharing services like Yahoo Flick or Google Picasa. When you known location, capture time and sharing license of the photo, all kind of fantastic services can be created, like Microsoft Photosynth.

Technically geotagging works by embedded GPS coordiates into the EXIF metadata of JPEG files.

N900 has geotagging as out of the box feature – no additional software needed. Just turn on it on in Camera application settings.

Also, you can retrofit your photos with geotagging information afterwards. You can do this by hand using labels and drag and drop in the most of photo sharing applications, like Google Picasa. Also there exist automated tools if you have relates GPS records available as GPX or KML file: checkout GPicSync. This is handy if you record your sports in eCoach and forgot to turn on geotagging in N900 camera. GPicSync also has a Google Maps export feature if you want to create custom maps for your friends or customers.

• An example map created with N900, eCoach,  GPicSync and Google Maps from one of the runs above

Sports tracking + geotagging = ?

I am still trying to figure out how to combine sports tracking and geotagging to something cool. Maybe something along the lines of urban exploration.

But in any case here are some of cities I have “collected” from my travels

• Budapest
• London
• Copenhagen
• New York (GPS was not exactly working well…)
• Oulu

PhoneGap ported on N900 (Maemo)

We have ported PhoneGap mobile application framework to the new Nokia N900 smartphone and its Maemo operating system. PhoneGap is a framework to build mobile applications easily with HTML and Javascript. With the new Maemo port PhoneGap platform support covers iPhone, Android, Nokia Series 60, Blackberry, Windows Mobile and Maemo, making it one of the most portable mobile application development solution available.

The demo application currently supports only Device Info and Accelerometer APIs. The porting work is still in its early stages and we suggest novice developers steer away from it. We’ll develop and maintain the work as long as we have client cases for mobile applications. If you are looking forward to port your commercial PhoneGap application to N900, please contact us.

PhoneGap port was done using QT and QWebView controller. The native shell source code is in C++. Build and packaging scripts are standard Makefiles. More technical information on the release notes page.

Code is available on GitHub.

So what’s cool about Maemo (compared to other PhoneGap platforms)?

Shortly: The openess of Maemo platforms enables developer innovation never seen before. There are zero artificial limitations chaining your imagination.

• No code signing whatsoever required
• No approval process to get your application distributed
• Very robust development tools and development environment. Hey, it’s Linux! The phone ships with X terminal built in.
• You have root access to the device if needed
• QWebView WebKit control which itself is open source – you can recompile from scratch and stick in the needed features
• The phone itself is rock solid. It beats iPhone 100-0.
• Maemo has very active open source community. You actually might get help when stuck, unlike with NDAs and other madness from some other vendors.

So what’s cool about PhoneGap (compared to other mobile application technologies)?

• Use low entry level HTML, CSS and Javascript technologies – even PHp coder can build his/her own mobile application! This is the most cost effective way to develop non-CPU-intensive applications.
• PhoneGap has the widest mobile platform support – the best medicine against fragmented mobile application markets
• You can always break out from the sandbox and use the native capabilities of the phone. This is something you cannot do with Flash Lite or Java ME.
• Very active community
• WebKit rendering engine enables CSS3 goodies and much more
• Easily convert your existing mobile site to an application

Installing Python Imaging Library (PIL) under virtualenv or buildout

I have greatly struggled to have PIL library support in isolated Python environments like virtualenv –no-site-packages.

For example, when installing Satchmo shop under virtualenv:

../bin/clonesatchmo.pyhe Python Imaging Library is not installed. Install from your distribution binaries.

../bin/clonesatchmo.py
The Python Imaging Library is not installed. Install from your distribution binaries.

Though it clearly is there, installed by easy_install PIL command:

ls ../lib/python2.5/site-packages/PIL-1.1.7-py2.5-linux-x86_64.egg
ArgImagePlugin.py	 ExifTags.py		  GimpGradientFile.pyc...

Does anyone know if this problem is with PIL itself, eggified PIL or something else?

In any case, there is an easy workaround: use system-wide PIL (sudo apt-get install python-imaging) and symlink PIL from your site-wide installation under the isolated Python environment:

(satchmo-py25)mulli% pwd
/srv/plone/mmaspecial/satchmo-py25/lib/python2.5/site-packages
(satchmo-py25)mulli% ln -s /usr/lib/python2.4/PIL .

Setting up multi-touch scrolling for Ubuntu 9.10 Karmic Koala Linux on Asus Eee 1005HA netbook

This post is specific to Asus Eee 1005HA netbook, but the technique explained here can be used on any computer having Synaptics touchpad.

Multi-touch gestures allow you to perform user interface actions by doing two finger gestures on touchpad. Apple introduced this feature on Macbooks and after you get used to it, it greatly enhances your web browsing on mouseless netbook. The most important gesture is scroll text by swiping the touchpad with two fingers.

Apple has also many patents related to the gestures so they are not enabled by default.

The real multi-finger touch support needs multi-finger aware (capacitive) touchpad. Most PC laptops are not equipped with one. Luckily some of the simple gestures, like two finger scrolling, can be emulated on normal pressure point sensitive touchpad via clever calculations and other tricks.

Note: Ubuntu HAL support for Synaptics seem to be broken. Only shell script at the end of the post will work. HAL options in FDI file are being ignored.

Setting up Synaptics driver

Type in terminal

gksudo gedit /etc/hal/fdi/policy/11-x11-synaptics.fdi

Create and save file with this content:

<?xml version="1.0" encoding="ISO-8859-1"?>
<deviceinfo version="0.2">
 <device>
   <match key="info.capabilities" contains="input.touchpad">
       <merge key="input.x11_driver" type="string">synaptics</merge>
       <merge key="input.x11_options.SHMConfig" type="string">On</merge>

       <merge key="input.x11_options.EmulateTwoFingerMinZ" type="string">40</merge>
       <merge key="input.x11_options.VertTwoFingerScroll" type="string">1</merge>
       <merge key="input.x11_options.HorizTwoFingerScroll" type="string">1</merge>
       <merge key="input.x11_options.TapButton1" type="string">1</merge>
       <merge key="input.x11_options.TapButton2" type="string">3</merge>  <!--two finger tap -> middle clieck(3) -->
       <merge key="input.x11_options.TapButton3" type="string">2</merge>  <!--three finger tap -> right click(2). almost impossible to click -->
   </match>
 </device>
</deviceinfo>

This allows us to use synclient utility to watch touchpad real-time data in console window.

Now restart X

sudo /etc/init.d/gdm restart

And open terminal again.

Type in command

synclient -m 100

And you should see data like this scrolling in the terminal:

129.355  2912 3469  59 1  4  0 0 0 0 0  00000000   0  0  0   0   0
 129.455  2952 3529  59 1  4  1 0 0 0 0  00000000   0  0  0   0   0
 time     x    y   z f  w  l r u d m     multi  gl gm gr gdx gdy
 129.555  3283 3516  60 1  4  1 0 0 0 0  00000000   0  0  0   0   0
 129.656  3928 3517  60 1  4  1 0 0 0 0  00000000   0  0  0   0   0
 129.756  4364 3637  60 1  4  1 0 0 0 0  00000000   0  0  0   0   0
 129.856  4020 3329  49 1  4  0 0 0 0 0  00000000   0  0  0   0   0
 129.956  3634 3122  58 1  4  0 0 0 0 0  00000000   0  0  0   0   0
 130.057  3320 2957  60 1  4  0 0 0 0 0  00000000   0  0  0   0   0
 130.157  2779 3312  61 1  4  0 0 0 0 0  00000000   0  0  0   0   0
 130.257  2557 3739  61 1  4  0 0 0 0 0  00000000   0  0  0   0   0
 130.358  2636 3485  39 1  4  0 0 0 0 0  00000000   0  0  0   0   0
 130.458  2659 3104  60 1  4  0 0 0 0 0  00000000   0  0  0   0   0
 130.558  2671 2988  60 1  4  0 0 0 0 0  00000000   0  0  0   0   0

f column tells the number of fingers. w is the touched area width. z is the pressure.

If you put two fingers on touchpad and you see value f=2 then your hardware has multi-touch aware touchpad. Unfortunately Asus Eee 1005HA doesn’t seem to have one

Emulation approach

Synaptics driver can emulate two-finger touch with the following conditions

• Touched area width exceeds certain threshold (min width)
• Touch pressure exceeds certain thresholds

When the conditions are met the driver thinks “Wow looks this guy is pressing us really hard. maybe he is using two fingers?” Note that touchpad values are touchpad specific and values applying for one model don’t work on another computer.

Synaptics driver settings are described here. Synaptic driver settings can be modified run-time using xinput command. Run synclient -m 100 in one terminal window and change threshold values in other until you find correct emulation parameters for your laptop. Below is my xinput tests. Test scrolling on Firefox and any long web page.

moo@huiskuttaja:~$ xinput set-int-prop "SynPS/2 Synaptics TouchPad" "Synaptics Two-Finger Width" 32 7
moo@huiskuttaja:~$ xinput set-int-prop "SynPS/2 Synaptics TouchPad" "Synaptics Two-Finger Pressure" 32 280
moo@huiskuttaja:~$ xinput set-int-prop "SynPS/2 Synaptics TouchPad" "Synaptics Two-Finger Width" 32 11
moo@huiskuttaja:~$ xinput set-int-prop "SynPS/2 Synaptics TouchPad" "Synaptics Two-Finger Pressure" 32 50
moo@huiskuttaja:~$

Looks like the following parameters are good for two finger emulation for Asus Eee 1005HA:

• Width: 8
• Pressure (Z): 10

You can also use command synclient -l to dump the current settings.

Below is the final script you need to run during log-in (see note about broken HAL at the beginning of the post):

#!/bin/sh

#!/bin/sh
#
# Use xinput --list-props "SynPS/2 Synaptics TouchPad" to extract data
#

# Set multi-touch emulation parameters
xinput set-int-prop "SynPS/2 Synaptics TouchPad" "Synaptics Two-Finger Pressure" 32 10
xinput set-int-prop "SynPS/2 Synaptics TouchPad" "Synaptics Two-Finger Width" 32 8
xinput set-int-prop "SynPS/2 Synaptics TouchPad" "Two-Finger Scrolling" 8 1
xinput set-int-prop "SynPS/2 Synaptics TouchPad" "Synaptics Two-Finger Scrolling" 8 1 1

# Disable edge scrolling
xinput set-int-prop "SynPS/2 Synaptics TouchPad" "Synaptics Edge Scrolling" 8 0 0 0 

# This will make cursor not to jump if you have two fingers on the touchpad and you list one
# (which you usually do after two-finger scrolling)
xinput set-int-prop "SynPS/2 Synaptics TouchPad" "Synaptics Jumpy Cursor Threshold" 32 110

Jumpy cursor after two finger scroll

When you do a two-finger scroll and lift your one finger before the other the mouse cursor/scrolling may jump. Synaptics driver does not seem to have an option to filter out this bad event. If anyone knows solution for this please comment.

Other resources

• Linux Multi-Touch is a project to collect set of Perl scripts related to multi-touch events. If you want to customize your multi-touch experience and you have some development insight, this is the place to begin.
• Notes about multi-touch emulation (see the last comment).
• Diagnosing multitouch support and editing HAL files.
• Related idea on Ubuntu Brainstorm.
• Synaptics X11 driver source. Synaptics driver project page. Looks like there is no one central figure behind it, except for maintaining, and big bunch of patches from distributors.

Cannot sort custom content item in Plone folder listing

Bug: Plone folder manual sorting does not move items even though you try all tricks. The first suspect would be a Javascript bug, but it isn’t.

It is bug 8161.

Your custom content meta_type must not contain spaces.

You can fix this on-line by editing meta type in portal_types in ZMI and remove all spaces from meta type name.

Subversion global-ignores and .egg-info in Python/Plone development

Subversion does a good job by ignoring most of build/temporary/unwanted files by default.

However, there is one exception still existing at least in Subversion 1.6: Python egg folders. All folders whose name ends up with .egg-info should not committed or considered in version controlling actions. your.package.name.egg-info folder is generated inside your Python egg source folder when you run setup.py / setuptools.

If you are working with Python source code eggs, add the following line to your ~/.subversion/config

global-ignores = *.o *.lo *.la #*# .*.rej *.rej .*~ *~ .#* .DS_Store *.egg-info *.pyc *.pyo .project .pydevproject

Otherwise development tools like Mr. Developer might get confused.