…just another hippy blog

“WALTHAM, Mass.-02 Nov 2006-Microsoft Corp. and Novell Inc. today announced a set of broad business and technical collaboration agreements to build, market and support a series of new solutions to make Novell and Microsoft® products work better together. The two companies also announced an agreement to provide each other’s customers with patent coverage for their respective products. These agreements will be in place until at least 2012. Under this new model, customers will realize unprecedented choice and flexibility through improved interoperability and manageability between Windows® and Linux.” (source: Novell Press Release)

This is some interesting news (yes, it’s no news any more, but I had no time to comment it earlier) - though I am not surprised about it. Reading the “Joint letter to the Open Source Community” on the Novell website, there are a few interesting points in the agreement:

1. Novell and Microsoft are releasing each company from the other’s patent portfolio. This means Microsoft won’t sue Novell for patent infringements and Novell won’t sue Microsoft.
2. Virtualization. The companies will collaborate to allow Microsoft being virtualised on Linux and vice versa. Note: Microsoft and Novell are already collaborating on the Xen project along with other top-level companies Intel, RedHat and AMD.
3. Office Open XML. Novell will be writing the code to allow OpenOffice to access the new Microsoft Office XML file format.
4. “Under the patent agreement, customers will receive coverage for Mono, Samba, and OpenOffice as well as .NET and Windows Server.”

A recent post published on ZDNet by Adrian Kingsley-Hughes, “The world just isn’t ready for Linux”, has cought my attention.

The author states that there are “areas Linux could improve on in order to transform itself from an operating system into a desktop platform”: gaming, software support, stop assuming that everyone using Linux (or who wants to use Linux) is a Linux expert, hardware support.
Let me point the spotlight to the words “Linux could improve” and express my personal opinion.
Read more

I feel like I have to say some words in support of Wikipedia after reading a few “stories” on Slashdot.

Number One: Wikipedia and Plagiarism

Short extract (Spo22a writes):

“Daniel Brandt found the examples of suspected plagiarism at Wikipedia using a program he created to run a few sentences from about 12,000 articles against Google Inc.’s search engine.”

In short: 142 Wikipedia articles out of 12000 (that’s about 1,83%) were suspected of plagiarism (this means they contained information coming from some other source).
Number Two: Long-Term Wikipedia Vandalism Exposed

Short extract (Daveydweeb writes):

“The accuracy of Wikipedia, the free online encyclopedia, came into question again when a long-standing article on ‘NPA personality theory’ was confirmed to be a hoax.”

In short: the article is about a not widely known personality theory invented by Anthony M. Benis. The suspect is that the article has been mainly written by Benis himself to promote his theory. The article has been proposed for deletion and a debate has been opened.

Number Three: Wikipedia Used To Spread Virus
Short extract (eldavojohn writes):

“The German Wikipedia has recently been used to launch a virus attack. Hackers posted a link to an all alleged fix for a new version of the blaster worm. Instead, it was a link to download malicious software.”

In short: some smart idiots (please don’t call them hackers unless you know the real meaning of this word) sent emails with a link to the fake fix in the hope that unsuspecting users would trust a Wikipedia link.

Read more

Jason Famularo has made a nice Google module to display one or a set of Flickr pictures. The problem is that I want to display one of my own images in a random fashion. The Flickr API has a flickr.photos.search method that allows you to retrieve the photos of a specific user. The problem is that you cannot retrieve them in a random order. There are the only parameters that allow you to change the photo order:

min_upload_date

Photos with an upload date greater than or equal to this value will be returned. The date should be in the form of a unix timestamp.

max_upload_date

Photos with an upload date less than or equal to this value will be returned. The date should be in the form of a unix timestamp.

min_taken_date

Photos with an taken date greater than or equal to this value will be returned. The date should be in the form of a mysql datetime.

max_taken_date

Photos with an taken date less than or equal to this value will be returned. The date should be in the form of a mysql datetime.

sort

Deafults to date-posted-desc. The possible values are: date-posted-asc, date-posted-desc, date-taken-asc, date-taken-desc, interestingness-desc, interestingness-asc, and relevance.

This means that I had to find a “hack”
Read more

Trolltech releases the third and last technical preview of their Qt bindings for Java. Lots of work has been done since the TP2 and lots of work still needs to be done.

This version is based on Qt 4.2 and thus it can benefit from the new features in Qt 4.2 (see my previous post). Some of the major changes include support for SVG graphics (this was added to Qt 4.1), the QGraphicsView (there’s a funny demo showing some mice ;), support for JavaDoc and “real” enums.

Some words about enums: C++ and Java enums are conceptually very different. Enums are used in different ways in Qt. Some examples are the QEvent::Type extensible enumerator and the or-combinable flags in QIODevice::OpenMode. Java flags can only be used as enumerators. TP3 solves this issue allowing to use enums with Qt even if the syntax is sometimes different.

The Eclipse plugin has been updated too, and lots of bugs have been fixed. I used it for some hours without Eclipse crashing! The Qt Jambi resource system still needs to be integrated into the Qt Jambi Designer (and it’s Eclipse integration). You can add resources using the Action Editor in TP3, but the JUIC would produce wrong prefixes and the files woun’t get loaded. Trolltech developer Eskil Blomfeldt has told me that this issue will be solved in a future release and we will be able to use the resource system with the Qt Jambi Designer too.

I was reading some posts on the Trolltech Qt-Interest archive and it seems that KDevelop developers are working on a Java language parser. This means we may be able to see a Qt Jambi plugin for KDevelop in (a not so near) future. Jakob Petsovits and Italian Roberto Raggi (a former student at my own University in Pisa should be the main developers for the Java-code.