Differences in How CRWebI Use Universes | Crystal Reports XI Official Guide

Differences in How CR/WebI Use Universes

Because both Crystal Reports and Web Intelligence can use universes, this discussion of the differences begins by restating what is the same:

Both report engines utilize the universe to generate SQL.
Both report engines update their SQL when newer universes are available.
Both report engines take advantage of semantic layer overrides.
Both report engines can have multiple result sets generated from multiple query panel requests.

The main difference between the Crystal and Web Intelligence report engines' use of the universe is the existence of the microcube engine in Web Intelligence. Because Crystal Reports does not have a microcube engine and uses a fundamentally different reporting paradigm, some SQL generated via the universe and accommodated by Web Intelligence will not be accommodated by Crystal Reports. For example, any single query panel request that requires multipass SQL will not be accommodated. This is because there is no microcube to reassemble the multiple resultsets into a single entity (data provider).

Tip

Again, you cannot overlook the tie between the data model and the expected usage. If you have a complex report with little to no interactivity that needs to be built off of a transactional model, this might be a better use for Crystal Reports without a universe. If you have a fairly decent data model that could possibly be used for analysis (even though reporting requirements don't currently state it), you can use either tool to write the reporteither Crystal Reports with or without the universe, or Web Intelligence with the universe. If you want a high level of user interactivity with the report (drilling, sorting, filtering, object swapping, and so forth), use Web Intelligence. If you want ad-hoc reporting with great performance, the ability to ask sophisticated questions, and a high level of interactivity, but do not have an appropriate model, no reporting tool can help youyou must model the data to meet the reporting requirements.