Lone Wolves

A short while ago Ty W posted an interesting question on StackOverflow. Apparently, most graphics software cannot scale images the right way. Usually it's hard to notice the flaw but the linked article does a great job of explaining the problem.

PHP's GD library suffers from the same issue, but Ty discovered that the sample PHP program provided with the article did not work on partially transparent images. After a couple of hours of fiddling I managed to get a working solution.

Apparently, the imagegammacorrect() function in PHP deals badly with images that have an alpha channel. I suspect that it tries to apply the same calculation to the alpha channel that it applies to the red, green and blue channels. To work around this, my solution splits the aplha channel from the original image. The alpha channel is resampled regularly while the red, green and blue channels are resampled using gamma correction.

One of the great advantages of the OpenDocument format is that it is simply a zip file. You can unzip it with any archiver and take a look at the contents, which is a set of XML documents and associated data. Many people are using this feature do create some nifty toolchains. Unzip, make some changes, zip it again and you have a new ODF document. Well… almost.

The OpenDocument Format specification has one little extra restriction when it comes to zip containers: The file called “mimetype” must be at the beginning of the zip file, it must be uncompressed and it must be stored without any additional file attributes. Unfortunately many developers seem to forget this. It is the number one cause of failed documents at Officeshots.org. If the mimetype file is not correctly zipped then it is not possible to programmatically detect the mimetype of the ODF file. And if the mimetype check fails, Officeshots (and possibly other applications) will refuse the document. This problem is compounded because virtually no ODF validator checks the zip container. They only check the contents. In this article I will show you how you can properly create ODF files using zip.

I have just released a new feature for Officeshots: The ODF anonymiser. The ODF Anonymiser tries to make your document completely anonymous while maintaining it's overall structure. All metadata is removed or cleaned. All text in the document is replaces with gibberish text that has approximately the same word length and word distribution. All images are replaced with placeholder images. All unknown content is removed.

The result of the anonymiser is a document that has the same general structure but with made-up contents. If your original document does not work in a certain application, the anonymised version of the document should fail in the same manner. By using the anonymiser you can test your private documents without exposing the contents to our rendering clients.

I am happy to announce an exciting new feature for Officeshots: Integrated ODF validators.

Every ODF document that is uploaded is run through several different ODF validators. If the converted documents are also ODF documents (when you are testing ODF round trips) then those results are also passed through these ODF validators.

The results of the validators are made available on the request overview, the individual result pages and inside the galleries. Galleries now not only show all attached documents but also all results and a summary of the validator results. This way it becomes really easy to see which documents failed.

“By the end of this book, You’ll be well on your way to becoming a Linux expert” is quite a bold claim for a book that is aimed at people who only have some familiarity with Windows and networking. “Pro Linux System Administration” by James Turnbull, Peter Lieverdink and Dennis Matotek aims to do precisely that and surprisingly, it largely succeeds. In its 1080 pages it explains how you can set up and configure multiple Linux servers to operate a small business network. Starting with basic Linux management and working up the stack through networking, e-mail and webservers you will end up with a pretty complete network that includes document management, groupware and disaster recovery.

The only downside of the book is that it becomes terser as it goes along. Part 1 and half of part 2 are quite thorough. They explain what you are doing, why you are doing it, differences between distributions and possible gotchas. But as the book moves up the application stack these explanations become shorter and in some cases amount to little more than an installation walk-through. I think it would have been better to focus on fewer alternative applications and dive deeper into those.

This article was originally posted at LXer Linux News.

A lot has been written recently about so called “Open Core” software ever since Andrew Lampitt coined the term back in August of 2008. Many analysts have been critical about it, such as Richard Hillesley from The H Open in his recent article “Open core, closed heart?”. Many are also very positive about it such as Matt Aslett from The 451 Group. However, I think that most them are missing the elephant in the room: Open core is not sustainable in the long term because it represents the worst of both worlds. Open core tries to find a middle ground between proprietary software and free software, but it reaps the benefits of neither and inherits the problems of both.

Let me show you by example. SugarCRM is one of the more popular open core software products available. The company offers the Community Edition for free under a GPLv3 license but also offers a Professional and Enterprise edition under a proprietary license. SugarCRM has been around since 2004 but it is already showing many signs of not being sustainable.

Digg this article:

I have finished setting up the internationalisation and localisation frameworks for Officeshots. If you want, you can now help to translate Officeshots to your own language. Translating Officeshots can be done through our Pootle installation.

At the moment there are almost no languages configured yet in Pootle. The reason is that the CakePHP framework on which Officeshots runs has a different locale structure than what Pootle expects. This means I need to add every language by hand. If you want to start working on a new language, please post to the Officeshots mailinglist and I will add the language to Pootle and to Officeshots.

One of the requirements for the upcoming public release of Officeshots.org is that all uploaded files are run through a virus scanner before they are made available. Picking a virus scanner for this job was easy. ClamAV is open source, well supported, actively maintained and comes pre-packaged for Debian Lenny which we use for the Officeshots servers. Finding a PHP library to interact with ClamAV proved harder though. The 3rd party library page for ClamAV points to two different libraries that provide PHP bindings for ClamAV but both appear to be dead and expunged from the internet. So, I created my own using the clamd TCP API, and because Officeshots is built using CakePHP I implemented it as a Cake plugin.

You can download the clamd-0.1.tar.gz plugin or check out the source from my Subversion repository with the following command:

Or you can browse the repository online. In the rest of this article I will show you how you can use this plugin.

When working on the beta of Officeshots.org I ran into an interesting problem with file type and MIME type detection of OpenDocument files. When a user uploads an ODF file to Officeshots I want to determine the MIME type myself using the PHP Fileinfo extension. Windows user who do not have any ODF supporting applications installed will report ODF files as application/zip which is of no use to me. In addition, a malicious user could attempt to upload an executable file and report the MIME type as ODF file.

On Linux, the PHP Fileinfo extension relies on the magic file that is provided by the file package. The magic file contains a series of tests that can determine the file type and MIME type of a file by its contents. I found out that the magic file is incomplete for OpenDocument files. Below I will show you what is wrong with the magic file and how you can fix it.

Update 2009-06-29: I have now also created a patch against the original upstream file-5.0.3.

CakePHP has rapidly been gaining mindshare as a powerful and easy to use MVC framework for PHP. Mimicking Ruby on Rails, it allows developers to quickly prototype and build database driven websites and web applications. With increased popularity books usually follow. “Practical CakePHP Projects” by Kai Chan and John Omokore is one such book. It is aimed at advanced PHP developers who have some experience with CakePHP and builds on books like “Beginning CakePHP” (Apress, 2008). The book promised to show how to build practical, real-world web applications using the CakePHP frameworks.

Unfortunately “Practical CakePHP Projects” only partially succeeds in that. It is refreshing to see how applications are built that are different from the proverbial “blog” or “store” example (though both are used in the first chapters), but I find myself disagreeing often with how these applications are built. The chosen solutions often seem to work against the framework instead of going with it.

This article was originally posted at LXer Linux News.