SOA has Indian origins

Purist would argue that SOA is not a new phenomenon that is taking up the hype-curve of IT, but an age-old principle – since the days of early packaged applications.

I propose a hypothesis that SOA’s origin can be traced to India. Rather, our dear babus of Indian Government were the main motive for SOA to come into lime-light.

After the burst of the dotcom bubble, there was not much of activity in Enterprise IT world. Sales dropped to an all time low, which gave a much needed breather to IT vendors to think of better solutions to the Enterprise world. One can recall that earlier generations of packaged application vendors adopted “silver bullet” approach when addressing the Enterprises.

This slack time gave them enough dope to do some research and provide a better (?) solution.

During the dotcom boom, many vendors opened their own shops (captive centers) in India. Every vendor worth their salt appointed one of their senior executives in India to overlook their Indian operations.

These senior executives when they went back to their homes (mostly in California) they were totally changed people. Their Indian experience gave them new insights into operations and this learning can be credited to be the founding principles for SOA.

Imagine acquiring 200 acres of land in Hyderabad (or any other metro) for setting up “campuses”; roughly the following are the steps that one needs to take, along with the approximate stipulated timelines. (Timelines are estimates – not actual)

Register a company in India:

1. Apply for a limited company with the Registrar of Companies. (RoC) {2 days}
   
2. Get clearance from RoC for the name. {1 week}
   
3. Get clearance from Dept. of Commerce and Trade for FOREX clearance – so as to get money for other operations etc. {3 weeks}
   
4. Join the STP or SEZ scheme. {4 weeks}
   

First to acquire land:

1. Identify the location – in or around the metro; usually outside the metro. (3 weeks)
   
2. Then talk to all the “patta” holders of the earmarked land and negotiate the price. (2 weeks)
   
3. Then you need to talk to a solicitor and get legal opinion about the authenticity of the ownership documents. (1 week)
   
4. Then get the land surveyed by the Revenue department, so as to map the details give in the ownership document with that of the official records. (1 week)
   
5. Register the land in the company’s name. (transfer of ownership) (1 day)
   
6. Apply for the conversion of the land type; say from agriculture to industrial or from barren to commercial etc. (1 week)
   

Start building the structure:

1. Get the building lay out ready and applies for HUDA approval. (HUDA is equivalent of any Urban Development Authority in Hyderabad) (1 week)
   
2. Incorporate the recommended changes. (1 week)
   

So technically the entire exercise of starting from scratch to getting the new building and to start production in India should not take more than 10 weeks, but in reality this would take at least 6 months. (In Singapore, with/without SOA it takes a mere 72 hrs to get things going.)

There are some unwritten protocols that need to be followed called the goonda-tax, where you need to give 10% of the total valuation of the project need to be paid to the local “political” leader.

Despite all these steps the senior executives could not get their stuff started, due to the underlying babudom.

It was then that they were introduced to the concept of SOA. Just approach the SOA vendor (in case of these senior executive, it was the local power-broker) where you have a single window to get all the things done within a small period of time. The entire “orchestration” of getting approvals from different departments would be easy and fast. The executives need not run between pillar and post to get their things happen, but be assured that the SOA-agent is getting things happen for them.

That senior executives were really amazed at the concept of agents here (India), be it the ROC, or the RTA or the Banks or rather at any other government agency. The simplicity of getting things done without going through the ordeal was something that stuck them with new set of thinking.

They went back to their country and then thought of something similar that can make the tedious tasks in Enterprise IT systems easy and thus started their research in “Indian” angle.

They named their research outcome as SOA (pronounced as SO-aAh!). Origin for that word is a story in itself: that was the first sound they all heard when they went to any Govt. Office in India. “saab soyah” – meaning “sir is asleep”.

Thus proves my theory. 

SOA and Women

There seems to be many similarities in the life-cycle of a home-maker with that of IT folks in Enterprise. Neither can ever achieve the nirvana, of less work and more output.

Of the many promises of SOA, the prominent one is the reduction of burden on developers, maintenance personnel, engineers et al, who are commonly clubbed under a banner category called IT folks in an Enterprise.

Let us look at this one particular promise from a different perspective, by comparing this SOA fever in today’s IT World with that of electrification of households in early 1900s.

Paraphrasing Nicholas Carr from his latest book, “The Big Switch” which provides a deep insight into the electrification fever of early 1900 and later we will look at the SOA fever of 2000s.

{Page 89}

The Utopian Rhetoric was not just a literary conceit; it proved to be a powerful marketing pitch for the manufacturers of electric appliances. General Electric was particularly adept at playing to people’s native optimism about technology. During the 1920s the decade in which the pace of wiring the American homes peaked, the company increased its yearly promotional expenditures from $2million to $12 million dollars. It devoted much of the money in installing in the public mind what is called “a positive electric consciousness” through a concerted program of magazine advertisements, pamphlets and presentations at schools and women’s clubs. Typical of the campaign was a booklet called ‘the home of a hundred comforts’ which described in flowery prose and futuristic illustrations how electric appliances would eliminate most household work, bestowing a life of peace and leisure on formerly harried home-makers. Having electricity in house the companies marketers proclaimed would be like having 10 home servants.

Whether conjured up for literary or commercial purposes the Utopian future never arrived. Cheap electricity brought great benefits to many people, but its effects purely played out as expected and not all of them were salubrious. Tracing the course of some of the most important of those affects through the first half of the last century reveals the complex interplay between technological and economic systems and their equally complex ways it exerts its influence over society.

Later in the book, he provides some insights which can be anti-thesis to the so called benefits.

{Page 99}

As it turned out, though, the electric iron was not quite the unalloyed blessing it first appeared to be. By making ironing “easier” the new appliance ended up producing a change in the prevailing social expectations about clothing. To appear respectable, men’s and women’s blouses and trousers had to be more frequently and meticulously pressed than was considered necessary before. Wrinkles became sign of sloth. Even children’s school clothes were expected to be neatly ironed. While women didn’t have to work hard to do their ironing they had to do more of it, more often, with more precision.

{Page 100}

A series of studies of the time women devoted to housework back up Cowan’s observation. Research undertaken between 1912 and 1914, before the widespread adoption of electric appliances, found that the average woman spent 56 hours a week on housework. Similar studies undertaken in 1925 and 1931, after electric appliances had become common, found that they were still spending between 50 and 60 hours a week on domestic chores. A 1965 study again found little change – women were spending on average 54.5 hours per week on housework. A more recent study, published in 2006 by National Bureau of Economic Research, also found that the hours housewives devoted to domestic work remained steady, at between 51 and 56 a week, in every decade from 1910s through the 1960s.

There might be many reasons for the no-change state of amount of work, while the primary reason was the push in the levels of cleanliness. An office employee might have been forgiven for crumpled dress before the arrival of electric pressing machine (Electric Iron). It is supposed to be easy to press cloths with the electric appliance hence it is mandatory for one to have pressed his/her cloths. Earlier carpets were cleaned once in a while (hence the term “Spring Cleaning”) as opposed to current once a week routine, if not once a day.

Due to this perceived notion of “ease of work” the workload remained repetitive due to the changes in the social norms; while expectations changed due to these widely canvassed “easy” perceptions.

Here, we can draw a comparison between the frenzy electrification and the todays frenzy SOAfication of Enterprise.

Will SOA not change the perception of Enterprise business folks towards IT folks?

Will not these so called “ease” of development, lesser “time to market” and flexible maintenance, create a new set of business norms? Will this ease not overburden the IT folks, else at least make them do repetitive tasks like cogs in machine working for the Enterprise will?

Before the advent of “packaged” applications, businesses depended on the multicolumn spreadsheets and also were tolerant of certain omissions, due to system limitations; compare the same with a packaged application, where everything is supposed to be out-of-the-box, yet the IT folks are found spending their energies in “customizing” the out-of-box features.

I would like to draw attention of the readers to one particular sentence in the above write-up from Nick Carr.

It devoted much of the money in installing in the public mind what is called “a positive electric consciousness” through a concerted program of magazine advertisements, pamphlets and presentations at schools and women’s clubs.

Doesn’t it sound threateningly similar to what our today’s magic quadrants and analyst reports talk of? The innumerable conferences, summits, road shows, webinars etc trying to instill in Enterprise minds the benefits of SOA and countless dollars spent by businesses in POCs to peek into the SOA world.

Can we draw safe conclusions that like the ease of electricity increased or rather did not affect the workload of homemakers; SOA too will not or cannot affect the burden of the IT, instead may increase the burden?

May be, future may tell… 

SofwareAG acquires webMethods

This is the press report

I said something similar here

Tool Wars ??

TIBCO announces to free up their BPM tool - Business Studio; just days after webMethods showcased their latest Designer™ during the Integration World 2006 event.

The EAI arena seems to be getting heated up with offers and counter offers; adding additional features, freebies etc.

A word about webM's BPM tool (Modeler™/Designer™)- It is now interoperable with Eclipse™ and the models (business processes) can be exported to other BPM tools in form of BPML – which was not the case earlier.

Btw, Savvion was the first one to give BPM tools for free, during Gartner’s summit on BPM sometime back in March.

How true...

Check the post by Pete Lacey, about S for Simple..

How many of us remember that 'S' in SOAP was really meant for Simple - and in reality how simple it is...

Good one...

SOA - webMethods

Till recently webMethods Inc provided two classes of integration tools, for B2B and EAI as separate product sets, which were mutually-exclusive. With the release of 6.x, a common platform for all the integration needs was provided.

Trading Networks™ (B2B tool) hosted in the Integration Server™ could directly link itself with Broker™ (message bus), which was not the case as before; thus helping orchestrate Models™ (Business processes) during runtime.

Version 6.x (v6.0.1 to be exact) laid the roots for SOA enablement, with a tool called ServiceNet and by default all the internal FLOW™/JAVA™ services were made available as web services. Of course, there were some issues with stability of v6.01 and v6.1.5 due to “composite” enterprise approach, which were stabilized by v6.5.

During Integration World 2006, hosted by webMethods, it was announced that Beta version of webMethods 7.0 would be released by Dec 15, while the GA is by mid-March. There was a major emphasis on the “Composite Enterprise Approach” with many tools performing focused-enterprise-tasks thrown in to the enterprise Fabric™. Tools like Optimize™ (BAM), Portal™ (B2C), Manager™ (BAM once again, but with heavy client) are all stabilized and rolled in to Fabric™.

The current way the Fabric™ is defined is very interesting -A weaver of all the Processes, Services, and Business Rules via web services into a Fabric for Enterprise. [As different threads when kneaded together gives a cloth (Fabric) of choice.]

The major themes of the event [Integration World 2006] apart from release of v7.x were SOA Governance and BPM.

Recently webMethods acquired Infravio, a SOA Registry and Governance tools company; adding the X-Broker and X-Registry features to its product suite from v7.x. For the starters X-Broker provides the intermediately/mediation services and policy enforcer in run time, while X-Registry provides the design time, change time governance. Obviously, the ServiceNet suite of products is deprecated and customers using the same are now automatically upgraded to Infravio suite.

Talking of SOA, security is always a concern/question. The good news from webMethods product suite is – there is support for WS-Security, SAML and X-509 tokens “out-of-box”. There was an interesting discussion during the focus group meeting as to how much of ‘out-of-box’ is their out-of-box. [Many can recollect that the migration tools for migrating EDI from Gentran/TLE etc were also out-of-box]

The Focus Group Discussion also provided an insight into what CIOs from the customers, SI representatives and Partners had to say about the SOA and Integration issues in webMethods context. Though there was nothing much to write home about, a glimpse at the most widely discussed topic during the discussion is worth mentioning here.

- What do you think of the greatest hindrance to SOA in your organization?

i. A majority of the CIOs felt that the buy-in from business was the most difficult one. Businesses have been spending on IT, only to bridge that “gap” between systems and business dynamics. Now all agreed that business is skeptical about the SOA hype-cycle.

ii. The ownership and governance issues were rated the next biggest hurdle. True. Who is going to own which services? Else again everyone ends up where they started, multiple service providers for same set of services. Governance is definitely an issue, which is not technical, but organizational, which definitely was beyond the CIO’s purview.

iii. The definition or the border lines to demarcate where services should stop being coarse grained, so that the existing infrastructure can be leveraged. The same question was also asked a decade ago, when Object-Oriented approach took the enterprise by storm. How much of logic to abstract, how much to define in code?

iv. Finally it was the tool-enablement. Though the majority of the tools in the enterprise ecosystem were not their respective current versions [meaning not SOA-enabled] everyone was optimistic about a turnaround soon.

In general CIOs were not in a mood to be the first movers and were eager to know if someone else has done something in that space. The group was eager to take back any implementer’s learning, for helping the approval process from business for their own SOA initiatives. Unfortunately there were none in the group, who had any real-time experience.

It was TCS who made some comments in discussion that their SOLAR framework was best for SOA enablement; but they too did not have any live enterprise to showcase, save for a few POCs here and there. There would be interesting times ahead for all industry observers watching how webMethods can live up to the promises made about SOA.

There was an informative tidbit when someone said Integration Server™, itself was an ESB and one need not invest in buying a third party tool.

When all FLOW™ or JAVA™ are by default web-services and these services can internally invoke any business process, business rule or perform data translation/transformation services. As these services are hosted in Integration Server™ - by standard definitions of ESB, IS™ becomes an ESB.

Moving further into composite enterprise, from v7.x there would not be any technical differences between human-dependant business flows and normal business flow. It may be recollected that in v6.x and before, there were two tools to handles these two variants of business flow. The human dependant business processes were called WorkFlow™ with a separate server to help the users to login and work on the staged business process at their levels, while the normal business processes were called Models™ hosted on PRT-module of Integration Server™. This invariably had to be designed in two different design tools called Modeler™ and WorkFlow Designer™.

The distinction would be dissolved with Designer™ 7.x of webMethods. Both these genres of business processes are now hosted in PRT-module for runtime and are designed in a single design time tool Designer™. This design time tool has an Eclipse™ interface. The good news stops here, as one still needs the Developer™ for developing FLOW™/JAVA™ services.

This meager would help development and implementation of the complex business processes in a single view and also to avoid redundancy on the pieces of code used.

One thing that stuck out the most was the fact that told over and over ‘40% of webMethods customers were SAP™ customers’. Also SAP™ was positioned well in the overall event, with a separate technical track session to showcase the way webMethods can blend itself into SAP™ based Enterprise. Someone there murmured that the future Integration World event may be a separate track sessions in Sapphire event.

[Continued…]

Promise to Post

Wil post the following in the following days...

1. Impressions about the event
2. Latest from webMethods
3. SOA - webM roadmap
4. Uncategorizable

Sorry about the delays - due to time constraints.