I recently got a new computer at work and needed to reinstall Cygwin. Here is my install procedure using the very nice cyg-apt script. It's like Debian's apt-get for Cygwin! Note, this install process assumes that I have all my init files backed up and can copy them over to my new machine.

1. Run the Cygwin setup program and install the default packages plus wget and python.
2. Set the Windows user environment variable HOME to c:\home\sofeng. Note, I tried it with forward slashes, but then it doesn't work in my Windows batch file.
3. Start the Cygwin bash shell
4. Change the cygdrive prefix:
   mount -s --change-cygdrive-prefix /
   If you don't have admin privileges:
   mount -u --change-cygdrive-prefix /
5. Copy over my init files and startup scripts:
   
   • ~/bin/startxwin.bat
     This is the Windows batch file that I use to start my Cygwin/X environment. It does the following:
     • Adds ~/bin to my PATH
     • cds to my HOME directory
     • Starts the ratpoison window manager with a urxvt terminal running screen
   • ~/.Xdefaults
     Sets the colors, font, and scrolling behavior for urxvt.
   • ~/.bashrc
     
     • Sets the PS1 (prompt), EDITOR, http_proxy, and other environment variables.
     • Sets a lot of aliases
     • Sets bash filename completion to be case insensitive
   • ~/.cyg-apt
     
   • ~/.emacs
     
   • ~/.emacs.d
     
   • ~/.inputrc
     
   • ~/.screenrc
     
   • /etc/ssmtp/ssmtp.conf
   • /var/cron/tabs/sofeng
6. Install cyg-apt

   $ wget http://www.lilypond.org/~janneke/software/cyg-apt
   $ chmod a+xr cyg-apt
   $ mv cyg-apt ~/bin

7. Install good stuff (note, if you don't already have ~/.cyg-apt, you will need to run cyg-apt setup first):
   

   $ cyg-apt install rsync	Best backup utility
   $ cyg-apt install screen	Terminal multiplexer
   $ cyg-apt install xorg-x11-base	X Windows
   $ cyg-apt install rxvt-unicode-X	Better Xterm
   $ cyg-apt install diffutils	diff
   $ cyg-apt install gcc-core	GCC C complier
   $ cyg-apt install make	GNU make
   $ cyg-apt install xorg-x11-devel	X header and library files for compiling ratpoison or other programs
   $ cyg-apt install readline	used for ratpoison
   $ cyg-apt install cron	so I can schedule my backups and updatedb
   $ cyg-apt install ssmtp	Sendmail replacement so cron can email me
   $ cyg-apt install emacs	Terminal version of emacs for quick editing jobs from the shell

Other notes:

• I thought that I needed to add c:/cygwin/bin to my Windows path in order to use grep in Emacs Windows. However, I later found out that this step is not needed. It is generally bad practice to add c:/cygwin/bin to the Windows path. Instead, I added the required code in my .emacs file. See http://cygwin.com/faq/faq-nochunks.html#faq.using.ntemacs for more information.
• I used to keep my http proxy server information in ~/.wgetrc, but now I store it in the http_proxy environment variable. (Note, it is lowercase and not uppercase. I don't know why this is.)

How to sleep for 5 seconds in python:

import time

time.sleep(5)

How to sleep for 0.5 seconds in python:

import time

time.sleep(0.5)

Documentation: time.sleep()

In the first installment of the Sample Django Project, I installed Django and created a project. In this installment, I will set up the SQLite database. At first, I thought I had to figure out what kind of data to put in the database, but in actuality, I can create an empty database and fill it in later. That is what I am doing. Here are the steps.

1. Install SQLite 3

   sofeng@tortoise:~$ sudo apt-get install sqlite3

2. Edit settings.py
   cd to the mysite directory created last time.

   sofeng@tortoise:~$ cd ~/Web/mysite

   Edit settings.py and change the following 2 lines

   DATABASE_ENGINE = 'sqlite3'
   DATABASE_NAME = '/home/sofeng/Web/mysite/mydb'

3. Create Django tables in the database

   sofeng@tortoise:~/Web/mysite$ python manage.py syncdb
   Creating table auth_message
   Creating table auth_group
   Creating table auth_user
   Creating table auth_permission
   Creating table django_content_type
   Creating table django_session
   Creating table django_site
   
   You just installed Django's auth system, which means you don't have any superusers defined.
   Would you like to create one now? (yes/no): yes
   Username (Leave blank to use 'sofeng'): 
   E-mail address: [email protected]
   Password: 
   Password (again): 
   Superuser created successfully.
   Installing index for auth.Message model
   Installing index for auth.Permission model
   Loading 'initial_data' fixtures...
   No fixtures found.

4. Take a look at the databases created

   sofeng@tortoise:~/Web/mysite$ sqlite3 mydb
   SQLite version 3.4.2
   Enter ".help" for instructions
   sqlite> .schema

   You should see a bunch of CREATE TABLE statements. If you get the following error message, it probably means you used sqlite instead of sqlite3.

   Unable to open database "mydb": file is encrypted or is not a database

Well that was pretty easy. Next time, we'll create some models and actually write some python code.

I mentioned in a previous post that I have an Excel spreadsheet containing a bunch of information about the development C code I'm working on. It is a large table showing which variables are input and output from which functions. The variable names are in the first column and the function names are in the last several columns of the first row. I use "i" or "o" to denote if a variable is an input to or an output from a particular function. I also have a few columns for the variable type and description. A simplified example is shown in the table below.

As shown above, the table also includes the variable descriptions, type, and scope. In my real, unsimplified Excel table, I also include function-related information such as the function description, and the function prototype. This additional information makes the table very messy. Try to imagine where you would put the function description and function prototype information in the table above. I came up with a kludge solution, but it involved repeated information, and rows and columns with inconsistent meaning. I was realizing the limitations of a 2-dimensional table.

I had been interested in learning about relational databases, so this seemed like a good opportunity. Relational databases imply the SQL query language-- relational databases that use SQL are about the only kind around. They include: Oracle, Microsoft SQL Server, PostgreSQL, MySQL, and SQLite among others. I chose SQLite for my task because it is lightweight (no separate server), free, popular, and fast for small databases. (See http://www.sqlite.org/different.html for other distinctive features.) Also Python 2.5 now includes an interface module, sqlite3, as part of the standard distribution. See http://docs.python.org/lib/module-sqlite3.html for the documentation.

Relational model design

I could create a relational database using my original table in the Excel spreadsheet. However, this organization is not in the spirit of relational database design. The [relational] approach is to create many smaller tables and relationships between those tables. For my example, I created a variable table, a function table, and a variable_function table to indicate the relationship between the variable and function tables. Though it looks more complicated than the original single table form, inputting this into a relational database, such as SQLite, will allow for more advanced manipulation of the data. It is also much more [conducive] to maintenance and expansion. For example, now there is a natural place to put the additional function-related information which was so clumsy in the original single table. To do this, I add desc and prototype columns to the function table. If I needed to add more information to about the different modules, I could create a separate module table and include the module_id as a column in the variable table.

SQLite Implementation (w/ Python)

To implement this database, all I needed was Python 2.5. If you have an older version of Python, you can install the pysqlite module. Here is the Python code to implement the database above.

import sqlite3

# data
VARIABLES = (
    (1, 'foo', 'ModuleExt', 'double', 'Description of foo'),
    (2, 'bar', 'ModuleExt', 'double', 'Description of bar'),
    (3, 'knark', 'Module1', 'int', 'Description of knark'),
    (4, 'wert', 'Module1', 'double', 'Description of wert'),
    (5, 'jib', 'Module1', 'double', 'Description of jib'),
    (6, 'laz', 'Module2', 'double', 'Description of laz'),
    (7, 'kew', 'Module2', 'double', 'Description of kew'),
)
FUNCTIONS = (
    (1, 'ExtSource'),
    (2, 'DoThis'),
    (3, 'CalcThis'),
    (4, 'CalcThat'),
    (5, 'ModifyStuff'),
    (6, 'ExtSink'),
)
VAR_FUNC = (
    (1, 1, 'output'),
    (1, 2, 'input'),
    (1, 4, 'input'),
    (2, 1, 'output'),
    (2, 2, 'input'),
    (2, 3, 'input'),
    (3, 2, 'input-output'),
    (3, 3, 'input'),
    (4, 2, 'output'),
    (4, 4, 'input'),
    (4, 5, 'input'),
    (5, 3, 'output'),
    (5, 5, 'input'),
    (6, 3, 'output'),
    (6, 5, 'output'),
    (6, 6, 'input'),
    (7, 4, 'output'),
    (7, 5, 'output'),
    (7, 6, 'input'),
)

# get connection and cursor objects
conn = sqlite3.connect('iodatabase')
c = conn.cursor()

# create tables
c.execute('''create table variable (
    id integer,
    name text,
    module text,
    type text,
    desc text
)''')
c.execute('''create table function (
    id integer,
    name text
)''')
c.execute('''create table var_func (
    variable_id integer,
    function_id integer,
    type text
)''')

# fill tables with data
for row in VARIABLES:
    c.execute('insert into variable values (?,?,?,?,?)', row)
for row in FUNCTIONS:
    c.execute('insert into function values (?,?)', row)
for row in VAR_FUNC:
    c.execute('insert into var_func values (?,?,?)', row)

Querying the Database

Now that I have created the database, I can get answers to interesting questions, such as What variables are output by CalcThis? Here is the Python/SQLite code to answer this question:

c.execute(''.join([
            'SELECT variable.name, variable.module, variable.type, variable.desc ',
            'FROM variable, var_func, function ',
            'WHERE variable.id=var_func.variable_id ', 
            'AND function.id=var_func.function_id ',
            'AND function.name="CalcThis" ',
            'AND var_func.type="output" ',
            ]))
FORMAT = '%-6s%-10s%-8s%-20s'
print FORMAT % ('name', 'module', 'type', 'desc')
print '-' * 44
for row in c:
    print FORMAT % row

name    module    type    desc
----------------------------------------------
jib     Module1   double  Description of jib
laz     Module2   double  Description of laz

c.execute(''.join([
            'SELECT function.name ',
            'FROM variable, var_func, function ',
            'WHERE variable.id=var_func.variable_id ', 
            'AND function.id=var_func.function_id ',
            'AND variable.name="wert" ',
            'AND var_func.type="input" ',
            ]))
print 'name'
print '------------'
for row in c:
    print '%s' % row

name
---------
CalcThat
ModifyStuff

lkup "" > mymap.txt

I plan to do a side project in Django just to get some web programming experience. I'm not sure exactly what the site will be yet. Maybe my very own blog software with extensive tagging capabilities and maybe revision control. I also have an idea for a site to specialize in comparison searching. E.g. "python vs. ruby", "mac vs. linux", "kde vs. gnome", etc. Or, if I find someone who needs a website, I might work someone else's project. As long as I get to choose the technology. Anyways, I plan to document my steps here. I've already installed Ubuntu. Here are my steps in creating my Django website.

Install django

$ sudo apt-get install python-django

Test install

To test the install, I typed import django in the python interpreter.

$ python
Python 2.5.1 (r251:54863, Oct  5 2007, 13:36:32) 
[GCC 4.1.3 20070929 (prerelease) (Ubuntu 4.1.2-16ubuntu2)] on linux2
Type "help", "copyright", "credits" or "license" for more information.
>>> import django
>>> 

Create a project

I tried to follow the Django tutorial and ran the django-admin.py command. However, I got the command not found error message.

$ django-admin.py startproject mysite
bash: django-admin.py: command not found

$ sudo ln -s /usr/lib/python-django/bin/django-admin.py /usr/local/bin/django-admin.py

$ django-admin.py startproject mysite
bash: /usr/local/bin/django-admin.py: Permission denied

$ sudo chmod +x /usr/lib/python-django/bin/django-admin.py

$ django-admin.py startproject mysite

Start the development web server

$ cd mysite
$ python manage.py runserver
Validating models...
0 errors found.

Django version 0.96, using settings 'mysite.settings'
Development server is running at http://127.0.0.1:8000/
Quit the server with CONTROL-C.

It worked!
Congratulations on your first Django-powered page.

I have Ubuntu 7.10 Gutsy Gibson (Gnome) and installed the emacs22 package. Copy and paste did not work from Emacs to the Firefox search box and from the Firefox location box to Emacs. I did get it working after a little Google searching. But it made me think back to Tim Bray's comments in Back to the Mac regarding the importance of a simple thing like copy and paste.

Here is the fix, which worked like a charm. I got this from the Ubuntu forums. Put the following two lines in your .emacs file.

(setq x-select-enable-clipboard t)
(setq interprogram-paste-function 'x-cut-buffer-or-selection-value)

Basic shortcuts:
F11: toggle full screen
CTRL+T: open new tab
CTRL+W: close current tab
CTRL+F: find on page
CTRL+PGUP: previous tab
CTRL+PGDOWN: next tab
CTRL+P: open print dialog
ALT+LEFTARROW: back
ALT+RIGHTARROW: forward
ALT+HOME: home

import os 
os.getcwd() 
'C:\\Python25'