The Ufdi Network

When I moved over from Dreamhost to Slicehost, one of the (minor) motivators was the option to use something different than MySQL. Don't get me wrong, though: MySQL is a nice system, fast, relatively memory-efficient and so on. But what drove and still drives me nuts is that not every storage engine within MySQL supports transactions. MyISAM can be as fast as it gets, if it lacks transaction support, I won't use it for at least some of my tables. And having to manually check that every time I write a new app really started to annoy me. But how to move a site like this, that is using Django, from one database system to another (or to be specific, PostgreSQL in this scenario) as simply as possible?

Note that this is just a rough description of the process I used. The principle should work in general, but you might face some additional difficulties that I didn't notice or haven't noticed yet. And don't forget to make a backup of your old data before doing anything described here.

The whole process is remarkably simple once you know what you have to do. Django helps a lot in this regard thanks to its fixture-system, that is luckily DBMS oblivious. Basically the whole move works like this:

1. Dump the database into a fixture using python manage.py dumpdata --indent=4 > dump.js
2. Change your settings module to point to a new PostgreSQL database (which you should create before the next step)
3. Run python manage.py syncdb to create the tables within your PostgreSQL database
4. Clean this database from the initial values with python manage.py sqlflush | psql yourdatabase
5. Run python manage.py loaddata dump.js
6. Make sure that the sequences in the new database are up to date

For step 5 I wrote a small script that goes through each sequence in the database (since this database is used for one Django site and one site only this works :P ) and resets its value to the highest id of the associated table:

Another problem that has to be solved before you do step 5 is that dumpdata has a small problem with BooleanFields in the models. If you're using any of those (you or any contrib module you're using), you have to do some cleaning up in the JSON dump. The problem is that the values of these BooleanFields are dumped as 0 or 1 instead of false or true, which confuses loaddata to no end. To make it easy for any processing script, I told dumpdata to nicely format the dump, which you can do with the --indent option, which comes in handy when trying to fix this small problem.

So I wrote a little Perl script (don't hurt me, please) that just goes over the dump and corrects it:

Naturally you'll have to correct the fieldnames for your own site So basically before loading the dump again, run something like perl convert.pl dump.json > dump2.json && mv dump2.json dump.json.

As you can see, the process is really pretty straight forward, and, so far, it seems to work just fine

Just a small update for those of you reading this blog’s feed: If you don’t want to see my German articles, there is something new for you: A feed that only includes English posts.

http://zerokspot.com/weblog/feed/en/

Das gleiche gilt für diejenigen unter euch, die nur meine deutschsprachigen Beiträge im Feed haben möchten:

http://zerokspot.com/weblog/feed/de/

Enjoy

Just a small update for those of you reading this blog???s feed: If you don???t want to see my German articles, there is something new for you: A feed that only includes English posts.

http://zerokspot.com/weblog/feed/en/

Das gleiche gilt f??r diejenigen unter euch, die nur meine deutschsprachigen Beitr??ge im Feed haben m??chten:

http://zerokspot.com/weblog/feed/de/

Enjoy

Think about what it would look like if a versioning system like git and a pastie had a while, and you'd probably came quite close what the folks behind the project hosting service Github let loose on the web just yesterday: Gist, a pastie that supports versioning and much more.

There are always some apps people start writing when they want to familiarize themselves with a web framework or some other web toolkit. The first thing that comes to mind here is a simple weblog. ... another is a paste service, or pastie for short. There are currently so many of these little apps out there, it's hard to get an overview, and definitely not worth anybody's time to even try. Most of them were developed around a certain community -- list dpaste for the Django community -- or a certain IRC network, but all of them share more or less the same featureset and usecase: Allow non-registered users to paste some code snippets to be shared with the community and later on forgotten.

Pastie + Git = Gist

So I was quite close to a yawn when I read the heading of this post on TechCrunch this noon. I guess the word "pastie" anywhere does this to me nowadays. Anyway, the post is about Gist, a new pastie service by GitHub, which has quite a nice twist to it, or actually 2:

1. It allows you to version control a paste and therefor also allows stuff like forking etc. using git
2. A paste isn't restricted to just one file, but can consist of multiple files

And since the whole service is (naturally ) build around Github's git-infrastructure, you can also checkout each paste and work offline as with any git repository and then commit your changes back (Well, you can work freely up to a certain point, but anyway).

I personally think that Gist offers some really nice ideas and additions to the whole pastie-idea, but at the same time I have to wonder, who will really use those new features. At first when I heard "versioning + pastie", I thought that this would be great for a site like djangosnippets or any other "real" snippets site (compared to a pure pastebin) where you actually produce real code. For something like this, Gist lacks a way to find such snippets, like tagging or simply a title for a paste. If Gist aims to just be a pastebin, then I'm really not sure, if the whole versioning thing isn't an overkill for the service, but perhaps someone else finds a totally cool usecase for it

Another question would be: Do your pastes affect your storage limit? Hmm....

Think about what it would look like if a versioning system like git and a pastie had a while, and you'd probably came quite close what the folks behind the project hosting service Github let loose on the web just yesterday: Gist, a pastie that supports versioning and much more.

There are always some apps people start writing when they want to familiarize themselves with a web framework or some other web toolkit. The first thing that comes to mind here is a simple weblog. ... another is a paste service, or pastie for short. There are currently so many of these little apps out there, it's hard to get an overview, and definitely not worth anybody's time to even try. Most of them were developed around a certain community -- list dpaste for the Django community -- or a certain IRC network, but all of them share more or less the same featureset and usecase: Allow non-registered users to paste some code snippets to be shared with the community and later on forgotten.

Pastie + Git = Gist

So I was quite close to a yawn when I read the heading of this post on TechCrunch this noon. I guess the word "pastie" anywhere does this to me nowadays. Anyway, the post is about Gist, a new pastie service by GitHub, which has quite a nice twist to it, or actually 2:

1. It allows you to version control a paste and therefor also allows stuff like forking etc. using git
2. A paste isn't restricted to just one file, but can consist of multiple files

And since the whole service is (naturally ) build around Github's git-infrastructure, you can also checkout each paste and work offline as with any git repository and then commit your changes back (Well, you can work freely up to a certain point, but anyway).

I personally think that Gist offers some really nice ideas and additions to the whole pastie-idea, but at the same time I have to wonder, who will really use those new features. At first when I heard "versioning + pastie", I thought that this would be great for a site like djangosnippets or any other "real" snippets site (compared to a pure pastebin) where you actually produce real code. For something like this, Gist lacks a way to find such snippets, like tagging or simply a title for a paste. If Gist aims to just be a pastebin, then I'm really not sure, if the whole versioning thing isn't an overkill for the service, but perhaps someone else finds a totally cool usecase for it

Another question would be: Do your pastes affect your storage limit? Hmm....

Here are some interesting Zend Webinars coming up in August:

Aug 13th, 9am PST
What’s New in Zend Framework 1.6?
Come join the Zend Framework team for an hour long jaunt through the new features added in Zend Framework 1.6. If you currently develop using Zend Framework then you won’t want to miss this one.

Aug 20th, 9am PST
Continuous Integration with PHPUnderControl Part 1
PHP Under Control is a Continuous Integration system that hooks in with PHPUnit and SVN. In this first part PHP expert Jesse Lesperance will discuss its features and advantages of Continuous Integration.

Aug 27th, 9am PST
Continuous Integration with PHPUnderControl Part 2
PHP Under Control is a Continuous Integration system that hooks in with PHPUnit and SVN. In this part PHP expert Jesse Lesperance will be creating Unit Tests with Studio for Eclipse and then running them in PHPUnderControl.

Sep. 3rd, 9am PST
Zend Framework and Dojo Integration
This past spring, the Zend Framework team announced a partnership with Dojo toolkit to provide out-of-the-box support for Rich Internet Applications.

via: devzone.zend.com

After nearly more than a year of development Brian Rosner merged the newforms-admin branch into trunk last night. This marks the last missing branch getting merged into trunk before the 1.0 release. Big kudos and thanks to Brian and everybody else involved in the development on this branch.

Another big change happened last night: newforms is now finally the "official" forms library for Django. At least it now resides directly within django.forms.

Both of these changes are backwards-incompatible, yet fully documented (1, 2). For everyone running on trunk (and who hasn't been preparing for this merge) this means some work esp. if you've messed around quite a lot with oldadmin. Also some of the styling in NFA looks a little bit different than before thanks to the move away from the old form manipulator modules which added some special classes to each form-element. If it annoys you, check out #5609.

After nearly more than a year of development Brian Rosner merged the newforms-admin branch into trunk last night. This marks the last missing branch getting merged into trunk before the 1.0 release. Big kudos and thanks to Brian and everybody else involved in the development on this branch.

Another big change happened last night: newforms is now finally the "official" forms library for Django. At least it now resides directly within django.forms.

Both of these changes are backwards-incompatible, yet fully documented (1, 2). For everyone running on trunk (and who hasn't been preparing for this merge) this means some work esp. if you've messed around quite a lot with oldadmin. Also some of the styling in NFA looks a little bit different than before thanks to the move away from the old form manipulator modules which added some special classes to each form-element. If it annoys you, check out #5609.

When dealing with multiple languages and internalization in PHP, some of the default functions in PHP end up mangling up the unicode characters in PHP. This is evident when you have a lot of funny looking characters coming up on your web page instead of the actual characters. Apart from setting the UTF-8 headers in your HTML page, you should be careful on which functions you use to handle your strings. There is an extensions called mbstring which you can install in PHP which gives you a set of functions which are unicode ( actually multibyte ) ready.

ASCII characters store each character in one byte. Unicode characters like UTF-8 use multiple bytes to handle the wider range of character sets. Some of the built in functions in PHP assume each character is only one byte and ends up breaking multibyte characters due to this assumtion.

One way to ensure that your content doesn’t get mangled up is to substitue the regular php functions in your code with the mbstring variety. To get the entire list of mbstring functions, head over to: http://php.net/manual/en/ref.mbstring.php

A few examples of the function mapping are:

EMail Function : Instead of using mail, you could use the mbstring function mb_send_mail
String Functions: strtoupper becomes mb_strtoupper; strlen becomes mb_strlen, substr becomes mb_substr and so on…

Now instead of going in and changing all your code to become multibyte ready, PHP gives you an easy way to overload the default functions with the mbstring variety.

You can set a value to mbstring.func_overload in php.ini. The value set for this function decides which functionality is overloaded by default with the mbstring variety:

• 1 - overloads the mail functions. So you don’t have to substitute mail with mb_send_mail in your code. The mail functuion it self will work like mb_send_mail if mbstring.func_overload is set to 1 in php.ini
• 2 - enables string functions overloading
• 4 - enables regular expression functions overloading
• 7 - enables mail, strings and regular expressions overloading

I will try not to spend too much time on this book here since I'm rather pissed right now. When "Warpath", which was released in March 2006, ended it was quite clear in what direction the next book would go: the mirror universe and the whole story behind Iliana Ghemor. This book does at least the later, but leaves the first one open to the next books ... and that's my problem with "Fearful Symmetry" by Olivia Woods. The book doesn't really progress the story. It answers basically all questions you could have had about Illiah, but the actual timeline progression is veeery minimal. I guess I couldn't even spoil anything here even if I tried.

The book is split into two parts starting on each side of the paperback. One for events in the life of Kira Nerys and one for the history behind the character of Iliana Ghemor. The first side of the book (137 pages) tells some stories from Kira's time during the Occupation that are relevant for the second part of the book, but, beyond that, doesn't progress the timeline at all.

The other part about Iliana Ghemor describes how she became an agent of the Obsidian Order and tells the story of her life up to the end of the Dominion War, which also puts her into a rather special relationship with Skrain Dukat. So, if you've read "Warpath", you will notice that there is still a whole between the end of this second part of the events taking place in "Warpath".

My main grief with this book is that it doesn't really progress the storyline, but just fills in some holes. This wouldn't be a problem if it had been part of the Mirror Universe miniseries as it probably was first intended to be (based on some earlier cover-shots), but at least according to the final cover it isn't. Some characters appearing to be a little bit out of ... character doesn't improve my impression of the book either. Here I especially mean Kira but also Vaughn and I'm not all that sure about Ro, either.

That said, the story itself isn't boring and the writing style itself isn't bad at all. On the other hand, the plot is just in basically all parts completely predictable (except for 2 events).

If you're dying to get to know the Iliana Ghemor character and find out all about her motivation (or at least a big chunk of it), this book is definitely for you. This background info will probably come in handy in the books to come, but as Jeff Ayers over at trekweb.com wrote ...

While Iliana Ghemor's story is interesting and vital to what I predict is coming next, her story has bad timing. If we didn't have two years between tales, it would have carried more resonance.

After two years, I expected to see a lot of focus on the mainline of the story, especially after the release of the first two books of the "Mirror Universe" mini-series. On this front I was heavily disappointed with the first side offering not all the much important (apart from about 3 paragraphs).

The next book, "The Soul Key" is again by Olivia Woods and I guess will continue the events of this book. I really hope the focus gets closer to the mainline, otherwise I'm quite sure that I will be disappointed again

I will try not to spend too much time on this book here since I'm rather pissed right now. When "Warpath", which was released in March 2006, ended it was quite clear in what direction the next book would go: the mirror universe and the whole story behind Iliana Ghemor. This book does at least the later, but leaves the first one open to the next books ... and that's my problem with "Fearful Symmetry" by Olivia Woods. The book doesn't really progress the story. It answers basically all questions you could have had about Illiah, but the actual timeline progression is veeery minimal. I guess I couldn't even spoil anything here even if I tried.

The book is split into two parts starting on each side of the paperback. One for events in the life of Kira Nerys and one for the history behind the character of Iliana Ghemor. The first side of the book (137 pages) tells some stories from Kira's time during the Occupation that are relevant for the second part of the book, but, beyond that, doesn't progress the timeline at all.

The other part about Iliana Ghemor describes how she became an agent of the Obsidian Order and tells the story of her life up to the end of the Dominion War, which also puts her into a rather special relationship with Skrain Dukat. So, if you've read "Warpath", you will notice that there is still a whole between the end of this second part of the events taking place in "Warpath".

My main grief with this book is that it doesn't really progress the storyline, but just fills in some holes. This wouldn't be a problem if it had been part of the Mirror Universe miniseries as it probably was first intended to be (based on some earlier cover-shots), but at least according to the final cover it isn't. Some characters appearing to be a little bit out of ... character doesn't improve my impression of the book either. Here I especially mean Kira but also Vaughn and I'm not all that sure about Ro, either.

That said, the story itself isn't boring and the writing style itself isn't bad at all. On the other hand, the plot is just in basically all parts completely predictable (except for 2 events).

If you're dying to get to know the Iliana Ghemor character and find out all about her motivation (or at least a big chunk of it), this book is definitely for you. This background info will probably come in handy in the books to come, but as Jeff Ayers over at trekweb.com wrote ...

While Iliana Ghemor's story is interesting and vital to what I predict is coming next, her story has bad timing. If we didn't have two years between tales, it would have carried more resonance.

After two years, I expected to see a lot of focus on the mainline of the story, especially after the release of the first two books of the "Mirror Universe" mini-series. On this front I was heavily disappointed with the first side offering not all the much important (apart from about 3 paragraphs).

The next book, "The Soul Key" is again by Olivia Woods and I guess will continue the events of this book. I really hope the focus gets closer to the mainline, otherwise I'm quite sure that I will be disappointed again

Here’s a how-to from AskApache.com which shows you how to improve the delivery of your Wordpress blog. The article oulines a few hacks to the WP-Cache plugin to improve the cachability of Wordpress. After the Wp-Cache hacks they go on to give you a few lines which you can add to your .htaccess file which

• Set the a future expires headers which Apache sends out to the browser while serving content
• Disabled ETag headers
• Removes last modified Header
• Adds cache control headers

These quick hacks ensures that the browser keeps the downloaded files ( htmls, css, images,..) and pulls them from cache next time to improve the speed at which your blog displays to the users.

Head over to the article and get started with your hacks: Hack WP-Cache for Huge Speed Increase

• Three tips to protect your WordPress installation
• How To Redirect Permalink Using .htaccess
• Super Cache Makes Your Wordpress Blog Super Fly
• WordPress Hacks: Moving a Static HTML Site to WordPress

Matt’s just announced the latest release of Wordpress:

I’m happy to announce that version 2.6 of WordPress.org is now available, almost a month ahead schedule. Version 2.6 “Tyner,” named for jazz pianist McCoy Tyner, contains a number of new features that make WordPress a more powerful CMS: you can now track changes to every post and page and easily post from wherever you are on the web, plus there are dozens of incremental improvements to the features introduced in version 2.5.

More details on the latest release at the Wordpress Blog: WordPress 2.6

Here’s a video which shows a quick tour of Wordpress 2.6:

I just got One-Theme installed (after reading about it at John Chow’s blog) and configured on this blog. It’s a cool theme which comes with it’s own admin section where you can change a lot of the theme settings, like selection of featured articles, the layout of the site, logo upload and more. They also have a auto-config for Adsense, where you just add your adsense account number in the admin, and the ads are up and running instantly.

I had a few problems getting on some the css and some of the page templates to play well with the plugins I had installed, notably the wp_codebox layout broke after the theme was installed yesterday, but with the excellent support from the One-theme team had got the problems sorted out and now, I guess all the kinks have been ironed out. Lemme know if you face any problems surfing this site.

With the help of this theme I’ve added a featured video on the home page where I plan to keep some good tech tutorial or talk videos (updated weekly or when I find some good stuff  )

• One Premium Theme to Rule Them All?
• Adsense Articles: Making Money With Google
• 20+ Must-Have WordPress 2.5 Compatible Plugins

Anthony Short’s got a great article which shares tips on how to get your CSS cross-browser compliant everytime.

Cross-browser compatibility is one of the most time consuming tasks for any web designer. We’ve seen many different articles over the net describing common problems and fixes. I’ve collated all the information I could find to create some coding conventions for ensuring that your site will work first time in every browser. There are some things you should consider for Safari and Firefox also, and IE isn’t always the culprit for your CSS woes.

He summarizes the article by saying:

1. Always use strict doctype and standards-compliant HTML/CSS
2. Always use a reset at the start of your css
3. Use -moz-opacity:0.99 on text elements to clean up rendering in Firefox, and text-shadow: #000 0 0 0 in Safari
4. Never resize images in the CSS or HTML
5. Check font rendering in every browser. Don’t use Lucida
6. Size text as a % in the body, and as em’s throughout
7. All layout divs that are floated should include display:inline and overflow:hidden
8. Containers should have overflow:auto and trigger hasLayout via a width or height
9. Don’t use any fancy CSS3 selectors
10. Don’t use transparent PNG’s unless you have loaded the alpha

Read the whole article: How to get Cross Browser Compatibility Every Time

• GhostWriter Mod for Unreal Tournament 3 Importing Videos into Unreal tutorial - Mod DB
  Importing Videos into Unreal tutorial

Pownce released last night the first alpha of their new desktop client. The previous, at least for me, had become unusable due to login problem and such, so I'm quite happy that there is finally an official -- yet still alpha -- replacement.

This new version thankfully comes with a simplified style and also embeds content like YouTube movies and the like instead of requiring users to once again go to the actual website. The client still uses Adobe AIR and is therefor available on Windows, MacOSX and Linux.

Remember when Ariel more or less suggested to use PownceMonkey instead of the official app (if you had some problems with it)? Back then I thought it was quite an interesting move. With the release of this first preview of the next iteration of the official app, the team behind the app has also become bigger. This new version seems not only to be the next official app, but also the next version of the back then suggested PownceMonkey as explained by Jon Rohan, author of PownceMonkey and now co-developer of the official app.

So far the app works quite well, although it has some weird rendering issue for instance with YouTube videos, where the video is scaled down when it's getting close to the border of the window.

YouTube videos get deformed

Also when you go to the "People" menu, you don't get all your contacts, but just the "first" 10 and no pagination.

So basically you get what you should expect when downloading an alpha. The core functionally is mostly there, but that's about it. That said, the app looks very promising and I guess you can expect a much more Pownce-like feel from it than from the previous version; especially since you get content-preview now at last. Previously the lack of the content preview that at least for me made Pownce so interesting more or less rendered the app pointless apart from it's notification aspect since you still had to either download the attached content (or follow the link if it wasn't an attached file) or jump to the website of the note.

From a developer standpoint I'm also thrilled to see, that the app is open-source and available on Google Code (bah,not on Github or Gitorious :-P )

Pownce released last night the first alpha of their new desktop client. The previous, at least for me, had become unusable due to login problem and such, so I'm quite happy that there is finally an official -- yet still alpha -- replacement.

This new version thankfully comes with a simplified style and also embeds content like YouTube movies and the like instead of requiring users to once again go to the actual website. The client still uses Adobe AIR and is therefor available on Windows, MacOSX and Linux.

Remember when Ariel more or less suggested to use PownceMonkey instead of the official app (if you had some problems with it)? Back then I thought it was quite an interesting move. With the release of this first preview of the next iteration of the official app, the team behind the app has also become bigger. This new version seems not only to be the next official app, but also the next version of the back then suggested PownceMonkey as explained by Jon Rohan, author of PownceMonkey and now co-developer of the official app.

So far the app works quite well, although it has some weird rendering issue for instance with YouTube videos, where the video is scaled down when it's getting close to the border of the window.

YouTube videos get deformed

Also when you go to the "People" menu, you don't get all your contacts, but just the "first" 10 and no pagination.

So basically you get what you should expect when downloading an alpha. The core functionally is mostly there, but that's about it. That said, the app looks very promising and I guess you can expect a much more Pownce-like feel from it than from the previous version; especially since you get content-preview now at last. Previously the lack of the content preview that at least for me made Pownce so interesting more or less rendered the app pointless apart from it's notification aspect since you still had to either download the attached content (or follow the link if it wasn't an attached file) or jump to the website of the note.

From a developer standpoint I'm also thrilled to see, that the app is open-source and available on Google Code (bah,not on Github or Gitorious :-P )

For the last 2 days I've now been hard at work getting some of the achievements trophies which are now available in some PS3 games since the last firmware update. And since the only game I have that supports trophies, I've been playing tons of Super Stardust HD again After getting the first 2 for finishing the first 2 levels, I noticed that for most of the trophies you have to get the expansion pack which costs 3.99 EUR. Since I wanted to get it anyway, that's not a big deal for me, but I still think its weird and a bad idea, to be honest, to make only about half of the trophies for the core game.

But apart from this issue, it will be a really hard decision from now on, for what system to get a game. Previously, it was quite simple: If the versions were the same except for the Xbox360 version having achievements, it would always have been the Xbox360 version I would have bought. This will probably make a nice 180 when I finally get an HDTV. Then I will have identical gaming-quality with both consoles but the PS3 will be far less noisy

The interface on both systems is pretty simple, although I prefer the XMB over the Xbox360 user interface. A nice addition to what Microsoft offers is a percentage indicator telling you, how many percent of the trophies you've already got. Apart from that, there isn't all that much of a difference in functionality between these two. In both you can easily compare your stats with your friends, see what achievements you've got and so on. In the XMB it seems a little bit more straight forward though, especially when you're the game you want to see your stats for.

I also like the whole tier system with bronze, silver, gold and platinum trophies indicating their difficulty ... if it's used appropriately. In Super Stardust HD there is for example a silver trophy for surviving more than 7 minutes in the Survival Mode. Good luck getting this one. When I checked the high-score list last night, only 12 people world-wide managed to survive more than 7 minutes... Should something like this be really just a silver trophy? Or is the platinum trophy not the only one with some special requirements?