How does eccenca compare to SMILA?
eccenca is built on top of the eclipse SMILA project
- adding more functionality such as an UI for administration and
- tying the crawling and processing processes more into the context of a search index
To understand the differences between eccenca and SMILA you need to know a little about its history:
The History of eccenca
At brox we have developed the Enterprise Information Framework (EIF) which covers the processes of
- indexing the data from sources and putting them into a search index
- searching the index data
EIF is also an integration framework, meaning that it lets you add implementations for
- information extraction from various sources (e.g. DBs, Files, Lotus Domino) and
- retrieval technology that provides the search index (e.g. Lucene, Fast, ...)
We have developed EIF from ground up adding more and more functionality as needs arose. Because this project started as early as 2000 some of its underlying technology doesn't meet today's needs anymore and we decided to co-found the eclipse SMILA project, bringing into it many of the ideas and features of EIF.
However there are certain limitations imposed on the scope and possibilities of SMILA that we feel limit its usefulness in many enterprise application contexts. As a consequence we strive to fill in those gaps and do this under the umbrella of eccenca.
SMILA does focus particularly on extraction and post processing of information but less on making this information searchable. In fact, SMILA's search side stems largely from EIF and has been stripped down and geared to support Lucene, only. What's missing is the support for integrating other search technologies and the associated framework for this.
With eccenca we focus again more on being able to plug in different search vendors - albeit Lucene is currently the only one available for free as part of the Community Edition.
In eccenca we also
- provide the tooling around SMILA and eccenca so that managing such a complex piece of software becomes possible
- offer a market place for anyone to participate and offer their own solutions of this great framework
Differences between SMILA and eccenca
eccenca is a pure extension of SMILA in the sense that we don't limit its functionality by the extensions that we add to it. Therefore, whatever you can do with SMILA you can also do with eccenca. However, the following limitations apply:
Version Compatibility
We base eccenca on distinctive versions of SMILA. At the moment this is M1. Since SMILA is still in incubation phase each release might break API compatibility so that newer versions are not likely to work out of the box with the eccenca bundles. This migration takes some time and as a consequence there is always some lag between eccenca and SMILA releases.
See our [road map] on when we plan to incorporate new versions.
Supported Feature
Not all features of SMILA are equally or fully supported by all eccenca tools. SMILA has a vast number of configs and turning knobs. For eccenca it is impossible to take them all into regard for its more specific use case and conventions - at least at the moment.
The following pages describes this in greater detail.
eccenca Conventions for Crawling, Processing and Index
