Strigi: new version, new webpage

kde-logo-official
Today strigi announced the release of the version 0.5.1. More information about strigi can now be found on a new strigi website.

The new website went online at the end of march. I’m not totally sure if it was announced and I just missed it because I was offline at that point or if it didn’t hit the masses.
In either way, the new website is a nice improvement over the previous one: although it differs from the typical other KDE 4 websites or from other KDE websites in general it comes along with a nice design, structured information and background information. Such things are important if you want to attract new folks (from developers till translators).

However, the new page does not feature news about the newest strigi version atm, but these information are available at the kde-apps.org entry. You also find download links there to get the new version.
The changelog is provided at kde-apps as well:

  • API cleanup because KDE4 is now using Strigi
  • new analyzer models that make it easier to write plugins for text, xml and binary files
  • multithreading support for indexing
  • inheritance in the field properties
  • better documentation
  • native support for microsoft office and pdf
  • removal of many dependencies

As you can see several of the important changes are deep inside of the program, but this is due to the fact that strigi is already used in the current KDE 4 development versions.
However, from a pure user perspective the native support of MS Office formats is nice – although I must admit that the file support section of the web page could be a bit more detailed, it looks a bit empty right now. For example, what is about kmail mails, thunderbird mails, akregator news, etc.?
What I still miss [1] is the inotify support: you can activate it as a compile option, but it is still considered experimental and is therefore no default option.

Besides the technical development the GUI was also revamped: it now features a more self explaining window without tabs but including small little helpers like a list of all indexed files. Still, the interface still feels a bit buggy: the “add directory” button simply doesn’t work for me in my local folder, and the list of indexed files includes suspiciously many files which would be matched by the file exclusion filter (hidden files, for example). And the GUI has a histogram field which doesn’t provide any information.

But since this is a part of KDE which is under heavy development and still has several months until it will be released as a stable KDE 4 application such things are ok. I’m looking forward to the final GUI for strigi – after all, beagle sets quite high standards in this regard, and I hope that strigi will show us something similar for KDE 4 🙂

[1]: I would like to have inotify for this reason: I don’t like it when my laptop is hit by any scheduled mass action. This happens with prelink often enough, and it is disturbing enough for my laptop. inotify is a much more elegant solution, even if it means I have a small daemon running in the background all the time which is running presumable anyway. And I get a lot of news every day, so I would need to run the daemon every day at least.

Apache at 56% – what is wrong?

Tux
The newest Netcraft Web server survey shows again a shrinking of Apaches market share. It is now at 56%, followed by Microsoft with more then 30%.

The current survey explains pretty clear that Apache’s loss of 2.86% is mostly due to a new rating system of the surveys: beginning with this May all pages hosted by Google are not longer rated as Apache solutions, but as GFE (Google Front End). These have a market share of 2.3%, leveling the real loss of Apache to “only” 0.56%.

Still, it is quite a lot – keep in mind that Apache had a market share of over 70% only two years ago. No one would have estimated that Apache would lose 15% in only two years. And even worse, the direction is also clear: Apache is only losing market share – nothing else! Why is that so? Has Apache lost its value?
Or is the community facing something similar to Firefox – but the other way around? No one would have guessed that Firefox would spread so unbelievable fast.

Of course, the possible reasons for this are manifold:
Is Microsoft doing a better marketing? Or has the code quality of Apache dropped somehow? Are there problems in the Apache management/core developers group? Is there a new trend Apache has missed? Or is it just the typical spread of the monopoly, making Apache slightly more unattractive on Windows servers than IIS (keep in mind that there are more Windows servers out there than Linux servers!)?
I have no idea if it is such a reason, or maybe something totally different. But I would like to know. And I would like to know why there is no reaction to this.

OpenSuse survey results published

suse-chameleon
The OpenSuse project published the results of a survey launched in mid February. Over 25k people took part in the survey giving a pretty good image of the users of OpenSuse and Linux-Distributions in general.

Several of the results of the survey are not very surprising. For example, most of the users are men (98%), have a fast Internet connection and use OpenSuse as their main operating system (roughly 50%, followed by Windows with roughly 30%). The most wanted feature is better hardware support (roughly 70%, but multiple answers could be given).

Other results, however, are quite impressive, and not what I would have guessed. For example, more than 70% said they use KDE. I wonder if the survey was announced in the same way at places like planet Gnome and planet KDE. Also surprising was that only 70% use a dual boot system. That makes 30% who do have a computer solely dedicated to Linux. This might be due to server usage or simply a second computer running Windows, but still, it is more than I would have expected. I always wonder if there are other people who are so “crazy” like me that they do not use any Windows system at all – although I have to admit that I work at Windows machines much more than I would like to.

The only thing I miss is that there is no question about software installation in general – there are questions related to the ease of the installation of OpenSuse, and you could judge the quantity of software packages with grades, but nothing was dedicated to the installation of additional software.

Still, it is a good survey, and it is very important to have such data available. I would like to see other distributions to pick up similar surveys – or, better said, I would like to see a general Linux user survey launched by the Linux Foundation, for example.