Once I was sure I was really sure that I was hearing what I was hearing, I got all confused about the date. Tuesday already? Where did the weekend go?

Please, Today, don't do that to me again.

I think I'm in love:

Anyone else up for it?

Not much to see, I'll admit, but I'm pleased - Matt and I made that happen.

Still, I'm not too worried. My eldest spent most of the weekend buried in Wesnoth.

Can it recommend a book that's not at all like The Da Vinci Code?

BTW, have you spotted The Shakespeare Secret? How blatantly fucking cynical can you get?

Sigh. I've been hearing about the death of the RDBMS for nearly twenty years. And the recurring problem with this thesis is not that the RDBMS is perfect, but that everything else is worse. Until someone comes up with an idea for a persistance mechanism that's better then an RDBMS across large parts of the problem space that the RDBMS solves, this is just idle hand-waving.

This effect had now moved on to the electric kettle. The new model in the office has multiple reservoirs so that you can fill it yet boil only enough water for one cup. Great idea. Trouble is, we had to have a session during our morning stand-up to demonstrate its use.

In The Atheist Delusion (see Godless evangelicals for a précis with comments enabled), John Grey says For Dawkins and Hitchens, Daniel Dennett and Martin Amis, Michel Onfray, Philip Pullman and others, religion in general is a poison that has fuelled violence and oppression throughout history, right up to the present day and These writers come from a generation schooled to think of religion as a throwback to an earlier stage of human development, which is bound to dwindle away as knowledge continues to increase. Now I think that they are certainly saying the former - and I agree with them. But I don't think that they all think that religion is bound to die out, only that it would be better it it did. Certainly there seems little evidence that the majority of people are becoming more rational, and if some of that irrationality is siphoned off into fringe beliefs like homeopathy and UFO belief, I think that's always going to be a limited group.

Yet Dawkins seems convinced that if it were not inculcated in schools and families, religion would die out. This is a view that has more in common with a certain type of fundamentalist theology than with Darwinian theory. So, Dawkins is a one dimensional character, is he? He's not allowed to hold opinions not directly related to Darwinian theory?

The idea of free will that informs liberal notions of personal autonomy is biblical in origin (think of the Genesis story). The belief that exercising free will is part of being human is a legacy of faith, and like most varieties of atheism today, Pullman's is a derivative of Christianity. I don't think anyone's saying that nothing that's ever come out of any religion was good. For example, Christianity's inclusivity was a major step forward for its time, with no built-in outgroup to despise. Christians have had to invent their own outgroups - mainly other Christians, for some reason that I'll never understand. Splitters!

Dawkins's "memetic theory of religion" is a classic example of the nonsense that is spawned when Darwinian thinking is applied outside its proper sphere. Unfortunately, the theory of memes is science only in the sense that Intelligent Design is science. Strictly speaking, it is not even a theory. This isn't entirely unfair, in so far as I've not heard of any serious work to verify memetic theory. But it seems quite falsifiable to me, in principle, so there's a least the germ of a theory there.

Wow - less than halfway through. More later. But once again, all this misses the crucial, central part of Dawkins' thesis, without which all the rest is minor, trivia: Religion, it's all just not true.

Back in September when I was last looking for a job in a moment of weakness I applied for a position at a global IT company that is rather fond of the colour blue. After the initial excitement of the application I waited.

And waited.

And forgot that I had applied to be honest, because I heard absolutely nothing. Until today, when reviewing the mails rejected by my spam filter I noticed this;

Dear Andrew

Thank you for your recent application to a large IT company*.

After careful consideration we regret to inform you that your skills and experience do not match
our current position requirements. If your skills match future vacancies we will contact you
again. Please let us know if your circumstances change, or if you do not wish us to continue to
hold your details.

Thank you again for your interest in a large IT company*.

Yours sincerely,
a large IT company* Australia Staffing Team

So by my calculations it has taken them five months to review my application and send a stock response. I was particularly impressed by the last line of the message which states;

This e-mail has been sent by a Service Machine. Please do not reply directly to this e-mail
using a Reply function.

Don’t you love the personal touch?

Update: I couldn’t resist and sent this reply;

Dear a large IT company*,

Thank you very much for taking five months to review my application. I do not wish you to retain
my details as I suspect that by the time you look at them again they will be woefully out of
date.

* Names have been changed to predict the gloriously incompetent.

Laptops, code review, show and tell etc allowed and encouraged.

We meet in the ground floor area next to P.J. O’Briens Pub internal entrance in the Grace Hotel, at the corner of York and King Streets in Sydney, New South Wales. See you there.

• A serious bug in the packaging script has been fixed and the distributed archive is now complete and valid.

Bug reports, patches and test cases to the usual email address please. Or just a message to let me know you are using the code would be nice.

Which has had an interesting effect on my approach to the software I use. Over the past couple of years I’ve been finding myself using either web based applications or command line scripts I wrote myself. What with having a desktop machine at work, a laptop and a home machine all running different operating systems the web browser and the command line are the easiest ways to synchronise my data via common applications.

As I’m now using the same machine at home and work I’ve started to drift back to o/s specific desktop applications. Instead of Google Reader I’ve switched to NetNewsWire Lite, I’ve replaced the WordPress text area with MarsEdit and I’m trying out OmniFocus instead of some hacked up scripts for my to do list.

I’m not sure what this says about me or my choice of technology, other than that I’ll get a frosty reception from the Google posse at the next Sydney Python meeting. But add this to the fact that I gave back the company supplied Blackberry and I think I must be bucking some sort of trend here. Is it just me or is anyone else fighting back the tide of web apps?

This Thursday, the 3rd of January, 2008 from 6:30pm, there will be a social gathering of the Sydney Python Users Group and any individuals interested in discussing Python, Web, Ruby, Perl etc.

Laptops, code review, show and tell etc allowed and encouraged.

We meet in the ground floor area next to P.J. O’Briens Pub internal entrance in the Grace Hotel, Cnr York and King Street, Sydney, New South Wales 2000

See you there.

1. If your database doesn’t do it automatically, add indexes to all of your primary key columns
2. Add indexes to all of your natural key columns
3. Add indexes to all of your foreign key columns

When primary or foreign keys are composites (i.e. contain more than one column) create an index containing all of the columns in the same order that they are defined in the corresponding constraint.

If none of this makes any sense and you are responsible for your application’s database give me a call, my consulting rates are very reasonable.

We meet in the ground floor area next to P.J. O’Briens Pub internal entrance in the Grace Hotel at the corner of York and King Street, Sydney, New South Wales 2000.

I’m going to do my best to be there, although my new job is keeping me somewhat busy. I should turn up even if it is only to increase the level of sartorial elegance and to show off my shiny new Macbook.

As part of setting it up I wanted to run some of my web pages locally and realised that PHP wasn’t enabled out of the box. Luckily it is really easy to set up. Just sign in to a terminal session as a user with admin privileges and;

$ sudo vi /etc/apache2/httpd.conf

and then uncomment line 114, which starts out looking like this;

#LoadModule php5_module       libexec/apache2/libphp5.so

Then restart Apache by de- and then re-selecting ‘Web Sharing’ under the ‘Sharing’ pane of System Preferences. You don’t need to do anything else because the nice people at Apple thoughtfully provide an appropriate configuration file that is included when you enable the PHP module.

• Support for PostgreSQL
• The code now largely conforms to PEP 8, thanks to excessive application of pylint.
• Packaging improved thanks to cheesecake

Bug reports, patches and test cases to the usual email address please. Or just a message to let me know you are using the code, that would be nice.

Expecting an update of every row in a 13 million row table (which is subject to frequent updates) to complete in one pass without timing out.

Of course, this wouldn’t be necessary if the system in question was running a version of PostgreSQL that was released after 2003. Damn those risk averse system administrators.

I should probably explain. In versions of PostgreSQL before release 8 you cannot increase the size of a column. Therefore to increase the size of one of the columns in our production database we have to add a new column, copy the contents of the old column to the new column, rename both columns and then drop the old column. Now that is what I call a database refactoring.

I shan't be posting any more entries to this blog but rest assured that the archives are staying. As part of this move comments have been closed on all of my existing weblog entries. If you have any problems with the new location or just fancy a chat send me an email. My address is andy47 at halfcooked.com.

It's a pleasure to install and work with, but I've had just one niggling problem. Oracle XE uses Application Express for administration and development, except that I couldn't get it to work.

Thanks to this forum thread (sorry, you'll need a login to read it) I just needed to perform a simple chmod 666 to my $ORACLE_HOME/network/log/listener.log file and everything started working fine.

Apparently the Oracle XE application express application is served through the good old fashioned Oracle database listener and it wasn't starting properly. Something as simple as the permissions on the listener log file were causing it to fail at start up, Something I'd probably have figured out myself if I'd bothered to look at the other log files.

Release 0.8.2 includes over 50 sample applications and tools to help users build applications in Python, including codeEditor, findfiles, and resourceEditor (layout editor).

New samples include a US-UK converter and a Sudoku solver. There are a new set of "convenience" functions to assist is creating pop-up menus and some commonly used custom dialogs (usage of these is demonstrated in the Sudoku sample, as well as in a new sample "helpful wrappers").

The tools/oneEditor directory contains a new version of the codeEditor, named tabcodeEditor.py that supports tabbed editor panes and the resource file popups from codeEditorR; once the code is more mature, it will replace the current codeEditor. The original codeEditor remains for now, but (unless problems show up) will likely be removed in the next release.

The tools/resourceEditor directory contains a new version of the resourceEditor, named multiresourceEditor.py which supports a number of operations on multiple components, such as Align, Distribute and Equalize, as well as visual tools to move components (Nudge) and to re-layer them. The other major change is that the Property Window has been redesigned to be larger, and display all properties at once, and allow editing of any of them immediately. The existing resourceEditor remains for now, though if the feedback on the newer layout is OK then the old version may be retired at the next release.

The tools/standaloneBuilder has been extensively revised, and now includes support for py2exe as well as pyInstaller.

PythonCard requires Python 2.3 or later and wxPython 2.5.2.8 or later.

You can download the latest release at:
http://sourceforge.net/project/showfiles.php?group_id=19015

Please be sure to look at the migration_guide.txt file in the docs directory if you are upgrading from a previous release. Since the package name has changed, you can continue to use the older PythonCardPrototype package simultaneously with the new PythonCard package, but you must upgrade to wxPython 2.5.2.8.

All the information you need about PythonCard can be found on the project web page at: http://pythoncard.sourceforge.net/

The installation instructions and walkthroughs are available on the main documentation page: http://pythoncard.sourceforge.net/documentation.html

For a list of most of the samples that have been built with PythonCard and screenshots of them in action go to: http://pythoncard.sourceforge.net/samples/samples.html

The kind people at SourceForge host the project:
http://sourceforge.net/projects/pythoncard/

If you want to get involved the main contact point is the Mailing list:
http://lists.sourceforge.net/lists/listinfo/pythoncard-users

My Powerbook is once more under the knife so I'm back playing around with Ubuntu on my work laptop.

Whilst perusing this Windows to Ubuntu transition guide I discovered that I can turn on user directories quite easily. Of course the instructions are a little misleading and the top Google hit, this Ubuntu forum thread doesn't quite work for me either.

Under Dapper and with Apache2 installed I had to;

$ sudo a2enmod userdir
$ sudo /etc/init.d/apache2 force-reload
$ cd
$ mkdir public_html

And now it's working like a charm. This change means that I can keep all of my files under my home directory and not pollute /var/www like I was doing.

Over the past month or so I've needed it again so I dusted that code off and made some enhancements. First I added support for MySQL and SQLite. As if that wasn't enough I then added an optional where clause parameter, so that you can dump some of the rows in the table. Then I added another optional parameter, this time a column list allowing you to only specify a subset of the columns in the table.

Then I just got carried away and introduced another function which accepts a valid SQL statement and dumps its results to a CSV file.

You can view the code here - csvDump.py or download it from here - csvDump.py.