Don't Try To Describe The Ocean If You've Never Seen It

Dont try to describe the ocean if youve never seen it
Dont ever forget that you just may wind up being wrong

--Jimmy Buffet, “Manana”

I read an article today on Fusion Applications written by Gavin Clarke for The Register, a United Kingdom-based website with the tag line “Biting The Hand That Feeds IT”. It's actaully dated December 11th, but I just saw it today (Dec. 12). You can read the entire article for yourself here.

There are more than a few inaccuracies in Mr. Clarke’s article. Rather than let the FUD spread throughout the Oracle user community, I’d like to address some of the more significant inaccuracies. My intent is not to “flame” Mr. Clarke, but only to make sure that everyone has the opportunity to work with accurate facts…

“Oracle will hit its goal of delivering Fusion applications next year in name only, with applications ready for testing but the full suite not due until 2009.” Along with several hundred other people attending the OAUG Fusion Council Panel at Oracle OpenWorld 2007, I heard Oracle’s Steve Miranda (VP for Fusion Applications) state that the first version of an integrated Fusion Applications Suite would be released in 2008. Steve also provided the caveat that the first release would not be a full functionality replacement for any current applications suite under Oracle’s Applications Unlimited umbrella (E-Business, PeopleSoft, JD Edwards, Siebel, etc.). What Steve described was an incremental production release in 2008.

“Fusion applications will be available during 2008 as an "early adopter beta suite" and the "full suite won't be until 2009", according the Oracle Applications Users Group (OAUG), which is close to the database giant's plans.” There have been rumors that the first release of Fusion Applications may come out through an early adopter program. In fact, it wouldn’t surprise me too much if things turned out this way - but that's strictly speculation on my part. At least for now, the latest official word is Steve Miranda's description of an incremental production release. Neither Oracle nor the OAUG has officially announced any "beta" . At this time, any rumor of an “early adopter beta suite” for Fusion Applications is just that: a rumor - as in no basis in fact. In the case of the OAUG, as the Co-Chair of the Fusion Council, I think I’d know if the OAUG made any statements along this line. Hmmm, let me check my email…voicemail…the OAUG website…Twitter feed…nope, came up dry, nothing at all.

“Coding on Fusion, at least, has begun, with Fusion apparently combining 80 per cent of the functionality from Oracle, PeopleSoft, JD Edwards and Siebel.” Oracle has told us for some time that Fusion Applications will be released incrementally. Will Fusion Applications eventually include 80 per cent of the functionality from the other Oracle applications suites? Probably; in fact, Fusion Applications will likely exceed the functionality provided by the others apps suites at some point. Will the first release include a combined 80 percent of the functionality from the other Oracle Apps suites? Not likely, in my opinion. Fusion Applications releases will be incremental and iterative, so the first release will not likely be too close to the finished product.

“All customizations for old versions of Oracle's E-Business Suite software will break and must be re-written using Java and XML.” What about the customizations, including bolt-on applications, already written in Java or using industry-standard Web Services? Answer – they’re likely to migrate pretty well (some of the interface architecture may need tweaking, for example). Ditto for moving Oracle Reports to BI Publisher (in fact, there’s already an Oracle tool for this migration available on a “managed release” basis. It’s the customizations written in PL/SQL, using mod_pl/sql, or Oracle Workflow that will be a challenge. Rather than stating that “All customizations…will break…”, a more accurate statement is that some of your customizations will need to be revised. Others will need to be migrated from legacy technology and tools. My advice is to catalogue your customizations, determine whether you still need them and, if you do, make plans for migrating those customizations to technology and tools supported within Fusion Applications.

“Applications such as Governance Risk and Compliance (GRC) manager are currently written on 11g, and do not use that 12g architecture as a jumping off point into Fusion.” Huh? 12g what? I thought the upgrade jump off points for Fusion Applications were determined by your current apps version. The closest thing I can come up with here is E-Business 11i (11.5.10.2) and E-Business R12…both of which are supported upgrade jump off points for Fusion Applications.

“Other so-called jumping off points to Fusion from Oracle's packaged software are Oracle E-Business Suite 10.2 and 11.5, PeopleSoft 8.8, 8.9 and 9.0, JD Edwards Enterprise 8.11 and 8.12, JD Edwards World A7.3, A8.1 and A 9.1. Jump off points for Siebel, Retek and iFlex have yet to be determined.” Oops, scratch my guess on the meaning of 12g in the last paragraph. Here are the apps jump off points for upgrading to Fusion Applications (referred to as “Other so-called jumping off points”. What other? How do you jump off to an upgrade if you don’t have some Oracle apps to start with?). So, how many users of the “Oracle E-Business Suite 10.2” do we have out there? Go ahead, raise your hands, don’t be shy… gee, no hands. That’s because the current support versions of the E-Business Suite are 11i and R12 – where in the heck did 10.2 come from?

In all fairness, I can't throw the baby out with the bathwater. The article does make two good points:

• “Developers…need to re-skill”. Absolutely true. Those of us restricted to writing PL/SQL will need to retool, learning Java or BPEL/Web Services if we want to continue developing in the Fusion Applications domain.
  

• “Another apparent problem with Fusion that's not being talked about by Oracle, is an increase in demands that Fusion's Java-based applications will place on server processing and server memory.” Well, this statement is half-right. Java-based applications do place more load on server CPUs and memory. I do, however, take issue with the statement that this is “…not being talked about by Oracle”. How much have we heard over the past year about the importance of distributed computing and virtualization to the overall Fusion architecture? At OpenWorld alone, there were over 60 presentations on grid and RAC.
  

One more point…I co-chair the OAUG Fusion Council with John Stouffer, who is quoted extensively in Mr. Clarke’s article. I’ve worked with John for many years in many different situations. Spreading FUD is just not his way. John is a reliable, accurate source of information and one heck of an Apps DBA. I have not had the opportunity to speak with John about Mr. Clarke's article yet, but I’m fairly sure he’s an unhappy fellow about the overall tone of the piece.

I hope this clears up some of the FUD before it spreads through the Oracle community like wildfire through a wheat field. Again, I’m not attempting to “flame” Mr. Clarke. I only wish that, in this particular case, he’d gotten the facts straight or at least gotten the product version numbers and names right.

I think I hear Jimmy singing Manana right now...

UPDATE: Just got off the phone with John Stouffer. Describing his mood as "unhappy" would be an understatement.