This is a bugfixing release - the previous versions deleted all files from the index directory when creating a new index (which caused problems when you mistakenly chosen My Documents as your index directory).
When you set the Seekafile.Web directory from the source package and try to access http://localhost/Seekafile.Web the following error appears:
Server Error in '/Seekafile.Web' Application. ------------------------------------------------------------------------------- Parser Error Description: An error occurred during the parsing of a resource required to service this request. Please review the following specific parse error details and modify your source file appropriately.
Parser Error Message: Could not load type 'Seekafile.Web.Global'.
The solution is simple: Because the source package doesn't contain the binaries you need to compile the project. Once you have all necessary assemblies in bin directory everything works well.
For your convenience, I've prepared a SeekafileServer-1.0.1-web.zip package that contains everything necessary for running the ASP.NET client search sample:
Full source code
Precompiled binaries
Index
Just configure it as a IIS virtual directory (e.g. Seekafile.Web) and you can try it at http://localhost/Seekafile.Web/.
The next release (1.1) will come on 15 July 2006. These are the new features:
Redesign of the plug-in system (plug-ins will return a Lucene.NET Document instead of a plain text).
IFilter support moved to a separate plug-in. (It used to be a default parser for everything that wasn't parsed successfuly. Now you don't have to use it for parsing if you don't need it.)
Management GUI for Seekafile Server service. Allows you to:
configure the watched directories, lock directory, index directory
manage plug-ins
check the index status
install/uninstall the Seekafile Server Windows service
release locks after a crash
Separate "Seekafile Server" (binaries and management GUI, probably and installer) and a "Seekafile SDK" (source code of the server, libraries and samples for creating plug-ins and client search applications)
Karel contributed a nice improvement of the Desktop Search Client. It is now released as version 1.0.1.
What's new there:
Supports result filters: file size, last modified date, directory (supports regular expressions), file name (supports regular expressions)Improved result listConverted to .NET Framework 2.0See the downloads page.
Screenshot:
Question from a user: On your installation documentation at http://www.seekafile.org/installation.html, there is this sample configuration config.xml file, which specifies the directories to be indexed and the path to the index files. I have the impression that even though the service indexes multiple directories, they can only be stored at a single location? What if I have two client
Question from a user: Why don't you add a spider/crawler to Seekafile Server?
Answer: Seekafile Server core is a file system watcher that watches changes that happen on a hard drive. It uses plug-ins to index additional files formats. Adding a spider would mean writing a totally new core that works in a different way. At this time I don't plan to write a crawler. However, if someone decides to
Question from a user: When do you upgrade the Seekafile Server to use DotLucene 1.4.3 or 1.9?
Answer: I will upgrade to 1.9 once it is production ready. So far I had problems with both 1.4.3 (concurrent write index access) and 1.9 (many strange exceptions, it's not stable yet).
The plug-in system used for documents parsing in Seekafile Indexing Server is very powerful - it allows you to heavily customize the indexing. Let's have an example:
You have a database where you store some kind of HTML documents (e.g. content management system). For most database systems, Seekafile Server (in combination with Lucene.NET) will give you much better performance, search results, and
Take a look at my article about building Seekafile Server on CodeProject: http://www.codeproject.com/useritems/seekafile.asp. Votes are welcome if you like it!
Features:
Incredibly fast search Incredibly fast searching powered by DotLucene (open-source search engine). It recognizes any change within a second. Flexibility The index is stored in DotLucene/Lucene 1.3+ compatible format. The index can be accessed directly from other applications (you can search even when the indexing is in progress). Access the index directly from any custom application
Question from a user: On your installation documentation at http://www.seekafile.org/installation.html, there is this sample configuration config.xml file, which specifies the directories to be indexed and the path to the index files. I have the impression that even though the service indexes multiple directories, they can only be stored at a single location? What if I have two client applications on the same machine, it is possible to have multiple IndexPath, one for each application?
Answer: If you need to maintain two indexes you will need to install two instances of Seekafile Server. I'm working on a multiple indexes support for a future version (1.2, see the roadmap) but the current version 1.0 can only handle one index.
Remember that:
A single Seekafile Server instance can handle multiple source directories
A single Seekafile Server instance can handle only on index
It is possible to have multiple Seekafile Server instances
If you want to install multiple Seekafile Server instances as services you need to change the service name in the source code (it's not possible to install two services with the same name)
The number of clients accessing the same index is not limited (it's even possible to access the index at the same time when the Seekafile Server updates the index)