View Road map to the future
Adding the possibility to make screenshot off all documents, not only images.
For some reason data converters, crawlers and security connectors are plugins, but the system for generating thumbnails from images and PDF files are not. It should be changes to allow for easier addition of thumbnail generators. For example it may be possible to use LibreOffice to generate thumbnails of traditional office documents like Word and PowerPoint.
The documentation should be Wikifyed to allow community contributions.
Drive for translations of the user interface
Setup a web based system that allows people to easily translate the user interface for us without having any knowledge about the ES or .po files. For example Simplepo.
Documenting exactly what is supported. Make a simple one page fact sheet that states what file formats, image formats, data sources etc. that is supported.
Long time strategic goals
Support extracting text from images and scanned PDF files.
Upgrade the underlying operating system
The ES is normally deployed together with a customized operating system that was derived from Fedora. The original Fedora version is no longer being maintained. We want to switch to a new CentOS version that has a life cycle of 10 years to free up time that we now are using on updating our operating system.
User space software only version
The ES currently dos many changes to the operating system it runs on, and can only be installed by using rpms. We want to make it less of a virtual appliance, and more a software suite that only runs in user space. Maybe stop supporting operating system configuration altogether and instead use a tool like Webmin on the appliances to setup the network and other system tasks.
Java interpreter for file converters and crawlers?
It is possible to develop file converters, crawlers and security connectors as plugins with Perl. Perl is supported natively because the ES has an integrated Perl interpreter (virtual machine). Maybe we should add a Java interpreter also to allow this types of plugins to be developed in Java. It may be a lot of existing file converters and crawlers that could be added quite easily.
Automatic distribution and clustering
Support for adding new nodes from the web based administration interface.
Implement HTTPS support
The ES should support serving request over HTTPS to make it more secure.
Fix all compiler warnings so the search core can be compiled with -Wall ( enable all warnings).
Currently there is no way of exporting a connector. Users have to copy the settings and source code from one ES to the next. A more user friendly way would be to have an export/import function that doesn’t require any technical knowledge.
Better push support
Support for more push options and JSON push.
Search results in JSON
Support for getting the search results in JSON, not only XML.
Allow for opt-out of path access
If the crawler supports it, the ES will look up the access rights for every file during search. This slows down the search if the source system is slow to answer. We should allow the user to disable this function if he wants to.
Allow administrator to set default search template
There is a “Client templates” section in the admin interface, but one has to craws a special link like http://demo.searchdaimon.com/webclient2/?tpl=myTemplate (not working) to use custom templates. It would be better if one instead just could set the custom on as default.