Blog

Created on Friday, 16 November 2007 17:47

Written by Larry Grey

Those who know who I am may find it wierd that I'm blogging this session as an attendee versus a presenter. This presentation was given by Colleen RinehartMurray, who I have a good history of working with (she took over the management of the reporting development team from me and allowed me to take on a pure product management role). You may also recognize her from past presentations where she and myself covered this exact session.

This session covered 3 things:

1. A review of how all the tools fit together.
2. A review of recent features made available in PeopleTools 8.48
3. A summary of planned items for PeopleTools 8.50

Tool Overview
Colleen discussed the different reporting tools, Crystal, XML Publisher, nVision, SQR, and Query and how they fit together. The slides did a pretty good job of covering that aspect, so I won't go into more detail on this part, except to say that Oracle is more aggressively positioning XML Publisher as an alternative to Crystal for new report development.

XML Publisher
There are several new features coming in XML Publisher. One thing to note is that many of these features are core XML publisher features, versus PeopleSoft-specific feature enablement. Here is the list:

• Subreports in XMLP. You can embed reports within other reports (from a template perspective). Examples of where you would want to do this is page headings and footings (there's a limitation in this covered in Jody Runstadler's Advanced Reporting Techniques session, which I will cover in my write-up for that session)
• Performance improvements for rowset and query feeding into it. Because PeopleSoft has to convert the data froma query and a rowset to an XML file on the server and then feed it into XML publisher, there could be potential performance issues. Development is working on an automated way of handling this in a manner similar to PS/Query.
• Email distribution. The delivered email distribution in process scheduler is currently only supported for nVision, Crystal, and SQR. This will be added for Query and XMLP (did I ever mention we have a product that already does this for any output type?)
• New versusions. They will add support for BI Publisher and 11g.

nVision
Started by making sure that the audience knew that nVision was not going away. For nVision, much of the discussion was on longer-term futures (8.50). The exception is for Office 2007 support.

• Not in tree criteria. This is something that we've seen a need for. It's mostly valuable when you want to capture data from of values that may not have been added to your tree (we approach it a little differently by proactively facilitating the addition of values to the tree as part of the maintenance process with this product instead of showing it after the fact)
• Multiple effective dated trees in same report. This allows organizations to compare the implications of tree changes side by side in a single report. Although you can see the changes pretty easily by exporting your trees to file and compare them there using tree mover, this allows you to see the dollar impact of the change.
• Query prompts. This is one of the features that I've heard a lot about over time. If you have a prompted query in an nVision report, those prompts can only be displayed in the windows nVision (not web). This will be an interesting feature if they really can get it in, because if they do it right, they may be able to add additional features as well. I'll put together a follow-up blog entry on the issues with this.
• Private reports in 2-tier. This is one of those things where there was a new feature added to the web that wasn't also added to the 2-tier nVision. You will now be able to secure your report requests on 2-tier as well.
• Excel 2007 adoptions of server performance improvements. This is something that I've had some discussions with Microsoft about and am very excited about. One of the biggest places where we had issues with nVision is stability and performance on the server. This is because Excel is a client tool where nVision automates it on a server. Excell 2007 has a server version that can be used very nicely with nVision, and they are adopting it. This will be a big win for anybody who moves to it (but my understanding is that Office 2007 is only supported on PeopleTools 8.49 or greater due to the file format changes imposed by microsoft. I'll have a lot more of that in my discussion on Jody Runstadtler's session.

Crystal Reports
Reiterated Oracle's commitment to continue providing Crystal as part of PeopleTools

• Includes Bobj XIr2.
• Mupltiple PeopleSoft instances sharing BOE server. When I scoped this feature for PeopleTools 8.48, this was one of the items that was very important to Business Objects (the business terms of this OEM were negotiated based on the fact that Business Objects would be able to re-sell additional licenses of Business Objects Enterprise for non-PeopleSoft applications). Much of the value of the up-sell is removed if there's one BOE environment for PeopleSoft and a different one for all other reporting, because that would cause users to have to sign into a different system to get one type of report than another.
• Note: per Colleen Murray, support for subreports has not been discontinued --> Discontinuation of Crystal Subreports. Now, this was news to me, and I'm curious as to the reasons why this is occurring. I'm sure it isn't an overt effort to steer customers to XML publisher from Crystal (although cynical folks could think that, based on the fact that subreports is a new feature in XMLP). I'll have to dig in more as to what technical issue cropped up for them.

Query
From a short-term perspective, much of the new functionality is in the area of extending the use of Query as a means of getting data out of PeopleSoft, specifically to use it as a means of quickly developing a web service.

• Subreport support. There wasn't much discussed in this area, but I'm pretty sure it's the work done in the Query Access Service (QAS) to allow multiple queries to be run concurrently through the same session.
• Connected query. Again, not a lot discussed, but again, I'm pretty sure this is the ability to use the query service to refresh data in excel without having to navigate to PeopleSoft to re-run it (support for pull as well as push).
• Extend web service enablement of query. I'm glad to see this one. There are a lot of situations where it would be valuable to take a query definition and add criteria on the fly and run it through a web service. Although I was planning a blog entry on doing this using a component interface, there were enough issues with the way the Query API works that I ended up not doing it.

Trees
On the tree side, there wasn't much to report in terms of new tree manager functionality, with the exception of web service enablement for trees. This would allow the Tree Classes to be exposed as web services. Also very valuable when you have two systems that you want to do lookups and updates across them (anybody have an HR Department tree and a Finance Department tree?). By having web services for this, you can do the lookups as you update one tree or the other. Very cool stuff.

Cube Manager
This was the big shocker for me. After many years of neglecting this product, I was sure it was going to die penniless and alone (although you wouldn't catch me writing that out formally). I've included the history of cube manager in this larger post on tree manager, in case you want to know why. Looking back on the Hyperion acquisition, it makes a lot more sense to keep cube manager alive.

• Support for new version of Essbase and PowerPlay. Yes. The versions being supported will finally be updated after 5 years. This will make many companies who have been staying on an old version of Essbase and PowerPlay because cube manager does what they want, and it just plain works (by the way, the kudos of this goes to a certain Brian Sparling, who created cube manager... Ever heard the name Sparling before??)
• Drop PeopleSoft star schema. The history of the PeopleSoft star schema is a very interesting one (but not one I can readily share in written form). Because it was created for EPM, but EPM is using Data Mart Builder to build its star schema, it makes sense that this feature should be dropped. Colleen did put to the audience that they should let her know if that feature is important to them, though.

Reporting Metadata
As mentioned above, much of the reporting metadata (queries and trees) are now going to be exposed as web services. This was reiterated.

• Using Query as a web service development tool. Talked about QAS (which was created for BOBJ enterprise)
• Exposing definition and designer capabilities for trees and queries. This facilitates support for OBIEE, sharing data, etc. It may also be interesting for Web Intelligence in Business Objects (with universes). I may have to talk to my friends at Bobj about this.

Reporting Infrastructure
As many folks who've gone to other sessions of mine will know, I spend a lot of time talking about the infrastructure to support reporting (metadata, security, scheduling, distribution). Colleen talked about what's coming in this as well.

• User Experience with respect to scheduling and monitoring reports. She mentioned that they're working on the ability to find reports that are older than 99 days. I'm sure that there's more in this area that they're not talking about. Did I mention that we have a product that fixes the user experience in this area?
• Making role-based security in report manager. The issue being fixed probably should have been a blog posting. Essentially, when you run a report to a role in process scheduler, it decomposes the role to its individual users at the time the report is run (take a look at the table PS_CDM_AUTH, you'll see that there is an OPRID field, but not a ROLE field). This means that if you move people in and out of roles, the reports associated with the role a person performs do not stay with the role.

Q&A
Although there were several questions on different partners and products, there is one that I want to make sure gets listed here. This was the question on XBRL support in PeopleSoft (which, for those who were at the session, may notice that I was the person who answered the question).

You can really think of XBRL as a standardized report format in XML for publishing financial results. In other words, it's a form (like your W-2 is a form). Because of this, the XBRL support by PeopleSoft is done by the applications groups (just like the support of W-2 is done by the Payroll development team). This support was provided in Financials 8.8 (and I remember the meetings with the developers to come up with the approach for doing this).

Here's the announcement of XBRL support: http://findarticles.com/p/articles/mi_m0EIN/is_2003_Sept_15/ai_107753879

Labels: 2007, Events

2007 | events