I am interested in hearing other people’s thoughts on advantages/disadvantages to using Crystal Reports XI and WebIntelligence XI? From a general point of view, I know that Crystal Reports is much more IT-centric (but really difficult to use for ad-hoc) and WebI is much more end-user centric (but not great for creating pixel-perfect reports). Some of my questions include:
What is the difference between a business view and universe?
Can you have one universe that can be used for both Crystal Reports and WebI? If so, what are the disadvantages to this approach?
What are other functionality considerations (e.g. drill down) of implementing Crystal vs. WebI?
Any other advantages/disadvantages would be great!
1.What is the difference between a business view and universe?
Business views are used in crystal reports to extract data from multiple data sources.Using that we can generate reports.
Universe contains the metadata information of a particular database.
2.Can you have one universe that can be used for both Crystal Reports and WebI? If so, what are the disadvantages to this approach?
In CRYSTALXI we can use universe to develop reports.
What are other functionality considerations (e.g. drill down) of implementing Crystal vs. WebI?
Crystalxi we have cascading prompt and dynamic parameter etc.
Crystal Reports is meant for canned reports (built by IT) distributed to a large user base or for pixel-perfect reports (forms etc.). WebI/Business Objects is meant for ad-hoc reports/analysis.
Thanks for the responses but that is not really what I was looking for. I know the obvious (Crystal Reports is IT-centric, pixel-perfect). I am more interested in hearing people’s opinions about things like creating a universe and using it for both Crystal Reports and WebI for XI.
So far, we’ve had trouble using Crystal to report off a universe. The query gets created with no problem, the report layout is no problem. When we preview the report in Crystal, we get prompted for a username and password directly to the backend database. i.e. Crystal appears to ignore the connection information on the universe. When we export the crystal report (based off universe) to InfoView, the report prompts the user for credentials to the backend database (regardless of whether integrated credentials are selected or not).
I’ve opened a case with support to find out how Crystal interacts with a universe.
This is an important point- Crystal Reports uses the Universe and Semantic Layer to generate the SQL that will be submitted to the database only. This is turned into a normal Crystal “SQL Command” which is then executed through Crystal Reports’ highly scalable data connectivity system- which is quite different from how Web Intelligence would execute the same query. Crystal Reports will require the database credentials to be provided specifically for each report, or stored in Enterprise, or configured for Single Sign On, regardless of how the Universe Connectivity options are set.
Also: Business Views provides a facility to Crystal Reports called “View Time Security” whereby the Report is executed against the database with a single hit, and then when different users view that scheduled instance it is filtered for the user security configured within the Business View, in memory without re-querying the database. This is a powerful and popular feature of Business Views, almost as popular as the configuration of heterogeneous data joins within the Data Foundation.
Well, I can see that may be a show-stopper for us for a given set of reports that we’re trying to convert from another tool (as an evaluation), which uses a special login for a special group of 15 users for just a couple of special data models/reports.
Snagoski: You explain the architecture of a connection to a universe as a data source very well. I wonder: Where did you get this information?. Is it 100% confirmed?. It looks like you simply looked at the interface to make your deductions. If XI simply uses the universe to determine the query required, then the connection string is a must (unless, like you say, the connection string can be stored togeteher with the universe object in the repository or BO Enterprise environment?). Is this the case?
there’s also a kb article about this (2017064). But the article implies that the prompting for username and password is a bug which was resolved in the XI SR2. Let me know your thoughts on this one.
I haven’t applied the patch yet. We’re running on UNIX, so the only patch available to me is for Crystal Reports XI itself (thick client). Has anyone had success only applying the patch to the thick client and not on the Enterprise level?
Thanks! I love it when I’m able to be Crystal Clear (cough) with my explanations.
As to your second question, I’m afraid I must be a bit orthogonal in my response about how I know, but I am in a position to speak with some authority about the architecture underneath this solution. It is definitely 100% confirmed.
The connection string is derived from the universe that builds the SQL (And any potentially applicable Overloads/Restriction Sets that might override the connection for a user/group) but it is stored local to the report and the CRDB interface makes the actual connection- not the Semantic Layer’s Connection Server.
Just one more thing. Does CR XI derive the connection string from the universe object definition in BO enterprise, or from the so called Universe Connection object (which appears stored separately in the BO Enterprise admin interface)?
Hmm, interesting question. it’s actually derived from the definition in the separately-defined Connection object, however the direction to which Connection object the connection info is fetched, comes from either the Universe or the Restriction Set (connection Overrides) object that is referenced by the report, and the users’ security credentials. Does that make sense?
(BOB member since 2004-07-16)
(BOB member since 2002-06-20)
(BOB member since 2004-12-13)
(BOB member since 2002-06-17)
(BOB member since 2002-11-22)