Monday, June 23, 2008

Oracle & BEA SOA perspective - Revisited

In a previous post in January I analyzed the SOA perspective of Oracle's BEA acquisition. I predicted that Oracle's future SOA solutions will be dominated by BEA's SOA solutions. Some comments presented different predictions from mine.
According to The Register article published on June 5th Oracle is splitting the BEA AquaLogic products line between Web 2.0 (originated from the Plumtree acquisition) and BPM (based on the Fuego acquisition). The Register also has learned from individuals close to the company that BEA's new owner Oracle is merging the AquaLogic and WebLogic professional service teams.

The AquaLogic Web 2.0 products probably will be overtook by Stellent Content Management which was acquired by Oracle on November 2006.
As described in previous post titled SOA & BPM, SOA and BPM are strongly related.
Another post tiled SOA and BPM: Too much Round Tripping? described an innovative product trying to extend this affinity beyond automatic BPEL based processes to include BPMN human based BPM.

Some comments on the AquaLogic line of products split suggest that AquaLogic BPM will be incorporated into Oracle Fusion Middleware and probably replace current Oracle offerings due to its superiority in human based BPM. (For example read Eric Roch's post).

The assumption is that the AquaLogic BPM will be incorporated into Fusion Middleware seems reasonable; however replacing current offerings is not the only scenario. Other possible scenarios:
  • Choice between two BPM solutions
This scenario is reasonable for the short time, but I would doubt if Oracle keep dual BPM offerings for the long time.
  • · AquaLogic BPM for development and integration of applications and Oracle BPM for Application Suites (e.g. Oracle Applications).
I still think that BEA's WebLogic will be finally Oracle's SOA Application Server.
The possible inclusion of AquaLogic in Fusion Middlewae together with merging AquaLogic and WebLogic professional service teams are first indicators supporting my prediction. Will other BEA"s SOA products included in Fusion Middleware keeping only the brand name and few components from the original Oracle infrastructure?

Monday, June 16, 2008

Modernization with SOA – The EDS way

Arye Amit and Eyal Kot presented EDS Modernization with SOA approach and practices in the 21st meeting of the SOA Forum of the Israeli Association of Information Processing on May 14th.
Arye is Business Development and Management Consulting of EDS Israel.
This post describes Arye's presentation about Business aspects and EDS approach to Modernization.

The following bullets summarize Arye's presentation:

Business Value of Modernization
  • Current Legacy systems include valuable assets; however the architecture is too rigid. Modernization is a key for transforming these assets to Agile solutions enabling the enterprise to respond to business changes and needs.
  • System Maintenance is about 80% of the IT Budget. Maintenance of old Legacy systems is expansive. After Modernization Maintenance costs will be reduced so higher percent could be allocated to Development and Innovation.
  • Omitting redundant systems developed by different departments.
  • Reducing Risk level due to diminished legacy skills.
  • High cost ineffective integration mechanisms are a barrier for fulfilling Business needs.
  • The goal is transforming the organization to a Responsive Organization

The Journey
  • Transition to SOA is part of The Journey towards Agile Platforms.
  • It includes Agile Enterprise (e.g. SOA), Agile Infrastructure (e.g. Virtualization) and Agile Applications (Standard, Open, Modular etc.).
  • Organizational factors such as Maturity should be considered
Modernization stages
  • · Definitions & planning: Mapping current assets, Roadmap etc.
  • · Managing the Journey: Business Architecture, Applications Architecture, performing the Modernization endeavor.
  • · Agile Enterprise (SOA, Business Processes, Vertical models etc.)
EDS's Approach

  •  Top Down
  •  Changing infrastructure to partners' infrastructure products (Microsoft, EMC, Oracle, SAP, Sun, Dell, Borland, CA, Cisco, Tibco etc.)
  •  Focus on Large Enterprises and on Outsourcing
  •  Gradual transition according to Roadmap and not Big Bang
  • The 7 Res: Re-learn (analyzing and understanding components, thinking of Reuse candidates), Re-host (migration to new platforms), Re-Factor (current code optimization), Re-interface (web based interface), Re-architect (SOA), Replace (replace Legacy applications by packages or new applications), Retire (applications retirement).

  • Critical Success Factors: Business Management commitment, Pragmatic approach (Best Practices, Methodologies), continuous plan adaptation, Budgeting next Modernization steps by savings from previous steps of the modernization endeavor.
Q & A
The following issues were raised by forum participants:
  • Organizational maturity.
EDS assess the maturity as part of the first steps based on a Maturity Model.
  • Organizational Culture changes and other organizational changes as part of the Modernization endeavor.
These changes are usually required during the transition process. The organization is maturing
gradually as described by the Maturity Model.

  • Partial Modernization
It is possible to implement part of the Modernization process as cited above. EDS defined and
is deploying autonomous services for partial implementations.
My Take
  • Modernization Value Proposition as presented by EDS is commonly presented by most SOA experts as SOA initiative Value Proposition.
  • Modernization as defined by EDS is not SOA only. Transition to SOA is a significant part of the Modernization endeavor in order to build a Responsive Organization and to get benefits of Agility.
  • A full implementation of EDS's methodology and practices is not a Panacea. It is a multi-years endeavor and high resources consuming. For some Enterprises it is a good choice but for others it is not.
  • It is not easy to justify this kind of endeavor. Some organizations should begin with addressing Pain Points by SOA project, which could be afterwards extended to a SOA initiative. Recently I read an Oracle SOA Case Study titled as: SOA by Accident: A successful point deployment of BPEL led to a SOA initiative.
  • Replace: is it necessary to replace platforms?
  • Replace: On which stage do you replace platforms: during transition to SOA or after completing the transition. The preferred approach is to postpone replacement after advanced SOA implementation stages. Afterwards a gradual transparent is a possible method. Sometimes maintaining current platforms could be inefficient, expensive or impossible.
  • EDS Modernization best fit could be large enterprises implementing the complete model and using previously EDS's services.

Saturday, June 7, 2008

Wikipedia: the Good the Bad and the Ugly

In The 7th post of Web 2.0 for Dummies I described Wikipedia. This post is dedicated to collaborators types named according to Sergio Leone's famous film from 1966 The Good the Bad and the Ugly .

The Good
Most of community collaborators are included in this type: people who are trying to contribute to community project.

Wikipedia's assumption as well as other Web 2.0 project assumption is that community members will contribute to the shared project according to the policy, guidelines and processes which were defined.
In many cases this assumption is true and the Good guys are enriching and improving Wikipedia's content. However, as the old proverb says: "The way to hell is paved with good intentions", sometimes despite of the good intentions the deliverables are not good enough. I will try to support my argument by examples.

Example 1: despite of a respectable reference
I had to clarify a specific Configuration Management aspect, as part of a consultancy project. I checked 4 references: three Web references I found by Google search and a Research Note by an analyst. The English Wikipedia definition of that issue was totally different from the other three sources.

It was crucial to find out which definition is right because of the practical implications.
It was difficult to decide. On the one hand I knew that the analyst is reliable (I participated in consulting assignment behalf of the same research and advisory company) and two other sources used an identical definition. 

On the other hand the reference on which the Wikipedia's article was based was an IEEE article. IEEE is a respectful information source. Unfortunately, I have no access to the IEEE article to check it.
After analyzing the data my conclusions were:
  1. The first definition (The analyst's and the two Web sources definition) is the right definition and the English Wikipedia's definition is wrong.
  2. Probably the IEEE article is correct as well and whoever created the Article in Wikipedia misinterpreted it.
Remember that most of Wikipedia's readers will not compare the content to other sources and analyze it like me.

Example 2: I know what I know Many years ago I learned the lesson cited in the header from an IT Guru. It is not a tautology. This statement implies humility: I do not know what I do not know. All of us know people that know what they do not know, in other words talk, write or even take decisions lacking the required knowledge. A frequent reason for SOA initiatives failures is basing it on internal skills of employees lacking SOA expertise and experience instead of using external expertise together with internal skills.

What drives community members to contribute without any payments?
The Psychological and Sociological motives are similar to motives in other societies or communities. The incentive could be Power. It could be Appreciation of other community members. It could be a feeling of being a part of a group.
Now imagine that someone motivated by Appreciation or/and Power is perceived as the Information Technology expert of his Wikipedia community. He wrote articles, checked and corrected articles and participated in votes and debates.

We know that usually modern experts knowledge is profound but restricted to very specific topic or role (of course there are exceptions to this rule).
This blog major topic is SOA, so as an example we will look at articles on SOA. Let us assume that the Wikipedia community expert's knowledge is roughly described by the following chart:
Topic Knowledge level
Development Environments Excellent
Operating Systems Good
Communication Partial
Hardware Partial
Integration Minimal

Architectures Minimal

SOA No knowledge
For illustrative purposes the chart includes only few IT areas.
The Good wrote valuable articles on Development Environments and few plausible articles on Operating Systems and DBMS and is considered as the IT expert.
Let us also assume that he did not learn yet the lesson: You know what you know.
The effects on SOA articles could be:
  • Poor articles written by the community IT expert lacking SOA expertise and thinking that he already acquired SOA knowledge for writing an encyclopedic article.
  • Flagging and changing of articles written by SOA experts who are not community members - His unconscious motivation is to keep is status as the IT expert and a SOA expert banish part of his status.
  • If a SOA expert will apply for community voting process for ensuring the quality of SOA articles he will probably fail, because most of the voters do not have any idea what SOA is. They will support there IT Good hero opinion.

The Bad
I do not know if the term "community members" is adequate for describing these characters. They intentionally destroy and alter articles written by others. They also write articles including deliberate errors.
The Bad may use simple mechanisms or very sophisticated Internet Bots. However, some articles quality deteriorates due to his "collaboration".
The period until restoration of previous version may vary according to the sophistication level of the Bad activity and the community auditing level and quality.

The Ugly
The Ugly does not destroy articles. He could be a respectful community member writing valuable articles, enriching other articles, voting, discussing etc.
However, he will not adhere to Wikipedia's neutrality principles, when the article is related to him. The motivation is personal benefit in the broader sense of the term.
The Ugly may be employed by a SOA vendor. He will keep this fact as a secret and will write articles on the company he is employed by, its products or articles on SOA technology which will be sophisticatedly biased towards his company's products.
The Ugly may be a politician or a supporter of a political party or a political view, biasing deliberately an article on that topic.
The Ugly may belong to a religious sect altering articles to reflect a good image of his sect.
Another Ugly could be an author praising his book in an article concealing the real identity of the article writer.

Wikipedia is a positive and valuable project but protect yourself from the Bad's the Ugly's and some of the Good's products, which could be hidden in any article.

Tuesday, June 3, 2008

Web 2.0 for Dummies – Part 7: Wikipedia

This post is another post in the "Web 2.0 for dummies" posts, based on my Web 2.0 presentation in a conference and my experience. After tasting Web 2.0 (part 2) and understanding what it is (part 4) and understanding Mashups and Virtual Realities, I will look at the case of Wikipedia.

What is Wikipedia?
Wikipedia is an Open Source (mainly an Open Content) community based encyclopedia. Its model is different from the traditional encyclopedia model based on books and from the Web 1.0 model based on loading a traditional encyclopedia to a Web site.
The difference between previous encyclopedia generations and Wikipedia could be summarized by the following bullets:

· Participation and Collaboration - Written by community members instead of experts. Writing is not the only pattern of participation: voting, editing, discussing are examples of other activities for expanding and improving the content.

· Multilingual – the community is composed of many sub-communities each of them working on different language. However, the common technological base enables cross participation between sub-communities by enabling
· Dynamic - participants may add, change, correct or delete content improving and adjusting the encyclopedic content. A similar change process in a traditional encyclopedia may take years instead of days.

· Hyperlinks – Encyclopedia Entries are related to each other by hyperlinks. Hyperlinks also serve as an easy way to access other related Web content sources.
Wiki is quick in Hawaiian language. This term describes the Agility of Wikipedia's content.

Economical and Technological model
The economical model is based upon donation and not on Advertising.
Technology is based on Open Source projects: Linux, PHP, MYSQL etc.
Major Wikipedia technology components include:
  • Wiki – a technology for enabling collaboration of multiple users to a Web page.
Version Management including restore to previous entry version is a crucial mechanism.
  • Web User Interface
  • Search Engines
  • Templates
  • Database Management and backups
· Dynamic and improving
Collaborators enhance and correct articles via a process similar to Open Source software code creation. An article is checked by many people so uncorrected errors rate can be low.
· Easy access to other related articles or data by hyperlinks
The hyperlinks hierarchy is composed of internal links, inter Wikipedia (cross languages), external (Web).

  • Multi-languages and Multi-Cultural

A local community can cover more deeply local topics (culture, people etc.) in comparison to to a globalized encyclopedia.
· Multi-media support


· Copyrights protection
Community members may copy protected text or pictures intentionally or unintentionally violating owners' rights to that content. As mentioned in previous post this challenge is common to many Web 2.0 projects.
Audits followed by deletion of that content written guidelines and disclaimer are the mechanisms for coping with this challenge

· Copyleft protection
Content usage is based on free content GFDL. Identifying and eliminating violations of it is a challenge.A commercial encyclopedia employee copying articles from Wikipedia is an example of violation of the GFDL.

· Systematic topics coverage
Experts creating a traditional encyclopedia will make sure that all significant items will be defined. In contradiction, in Wikipedia community members choose whatever articles to cover. The result could be missing important items as well as articles on negligible subjects. Communities try to address this issue by initiatives for mapping relevant subjects and terms in a specific topic and writing articles on them. Another issue relating to non-systematic coverage is related to the evolutionary nature of Wikipedia. Even if a topic is covered systematically in a specific time only subset of articles will be available.

· Content quality
Content quality is nonuniform: Some articles are excellent while other articles level is not good enough. The reasons for not good enough articles vary: intentional distortion, lack of expertise by the authors, lack of time for article evolution etc. Auditing, Deleting and restoration of previous version are mechanisms for addressing this challenge

Wikipedia is a successful and valuable Web 2.0 initiative. However, one should be careful because some of the content will not be adequate. The real problem in using Wikipedia content, as well as other Web content is to distinguish between excellent content and inappropriate content.

About Me

Share on Facebook