23 April 2009

Siebel: URL Tricks

Most of us use the URL to the Siebel application as just that - an address to reach Siebel functionality. But there are a few tricks that could make life easy for a few developers. I don't know of many, but will certainly start out with a couple of them I have used extensively in the past. Do you use any other instructions that may ease everyone's life?

  • Refresh browser script on the webserver: There are a lot of jittery admins out there who will not give you rights to restart the web server box. But at the same time, you don't seem to quash the bugs hard enough and you need more than a few SRF replacements. What do you do? Type this in the browser: http://host:port/application/start.swe?SWECmd=UpdateWebImages&SWEPassword=WebUpdateProtectionKey where: host & port = webserver address as in the URL used to access Siebel application; application = Siebel application identifier - sales_enu, media_enu etc); WebUpdateProtectionKey = Unencrypted web update protection key - defined in eapps.cfg using WebUpdatePassword parameter
  • Execute a few commands for quick and dirty debugging. For example: javascript:alert(theApplication().GetProfileAttr("ApplicationName"));
  • Test your HTTP post/get interface by directly providing the input XML in the address bar. For example: http://host:port/eai_enu/start.swe?SWEExtSource=myService&SWEExtCmd=Execute&UserName=&Password=&SWEExtData=<url encoded request XML>

Making Sense of Web 2.0 Data

Sure, the organization is a trend setter and religiously adopts all web 2.0 technologies and the management talks about it all the time in its internal and external blog postings, tweets and *gasp* facebook. But again when sales or marketing actually look at all the new channels, how exactly do they make sense out of it? How does anyone make sense off the unstructured data that lies out there, but is VERY important for the ecosystem to utilize the next gen of communication tools.

There have been a number of text analyser tools, tracking tools for all kind of "social" activities and there is a new guy on the block - Webtrends Social Measurement integrated with Radian6. Using this tool organizations can track all kind of Web 2.0 chatter including blogs, forums, alerts, feeds and what not. This should facilitate easy monitoring of what is going on about the company and at the same time provide a quick response or a solution where needed. It always helps to keep customers engaged in conversation..!

Read more


18 April 2009

CRM and Cloud Computing

For the past few months now, I have been following developments in the cloud and those are all cheery and bright. CRM being one of the not-strictly-in-house applications, is one of the prime candidates for "clouding" and obviously those of us who depend on CRM for a living are going to look at the developments closely. I am of the opinion that for better or for worse, moving services to the cloud is where we are headed. Software as a Service (SaaS) has already proved its mettle and has not limited itself to SMBs. However for those of the companies who just want their application but not the hassles and for those who may want independence to scale hardware and infrastructure as they wish, moving services to the cloud is an option. And that option has been picking up lot of heat with Amazon & its cloud services, Microsoft & Azure and so on & so forth.

With the above as backdrop what I really wanted to discuss is the recent McKinsey report, which has been interpreted at many a place as "cloud services are ineffective". In many ways, I see this hype/anti-hype as similar to that generated when initial SaaS discussions were on. SaaS came and is here to stay. But that did not make the current software market vanish (though I understand Mr. Benioff has been trying to do just that). Though the dust is not all settled today, we still see a lot of software vendors out there who are making a profit (with a decline on the cards) and at the same time there is a aggresive growth in SaaS market. With the ever lessening costs of SaaS, ever increasing competition and the ease & sophistication of customization options, we might see organizations moving to SaaS over a period of time - nothing overnight. Who knows SaaS vendors may unveil even better strategies to capture new segments by addressing many of the critical issues that face the customer today - yes, including data security concerns.

I see a similar analogy in the cloud services market that will include more than just the one-fits-all software. No data centers are going to move to cloud over night, but will be a gradual process. From where I stand, it makes business sense for the provider to put together a bunch of options and let the users use remote servers and services without the hassle of maintaining one. This has to be cost effective and can result in better resource utilization in the long run. But, not all organizations will be able to do it - may it be geographic constraints or plain "sense of ownership" thought-process. The McKinsey report does deal with such questions and the observations are simple:
  • Cloud reduces upfront costs
  • Almost immediate access to servers
  • Smooth scaling up of services on demand
Cloud services already has made a dent and does make a lot of sense for the SMBs but large organizations would be better off by not investing too much too quickly. Hosting your services in-house costs lot cheaper than putting them on Amazon. New technologies tend to garner lot of interest and that sometimes over-cooks it - too much hype will not automatically translate into benefits. I for one sincerely believe cloud computing will be mainstream and will define the standard in not-too-distant future. Ultimately what will work is the same mantra as IT outsourcing - "We want to concentrate on what we do best and let the IT part be handled by specialists". Here we will have organizations who want to be agile in their response to market and leave the infrastructure & software to be maintained elsewhere. What do you think?


Back for good

I am back after a long time, had to attend to some serious issues that would have taken away my source of income. And this is not what you want to do in these times of recession.

I have seen a couple of more projects hence and still wonder what makes project teams tick? There are a few people around who actually like what they do for living. Day by day, project by project you see the same mistakes repeated often. The same review comments given out by Oracle and the flurry of fixes. Unless you have a tight grip on ALL configuration objects or have someone interested do it, you are pretty sure to go through the 25 pager that Oracle ES doles out for minor defects in 25 objects!

Anyway, here are a few must-do's in all Siebel projects:
  1. Keep Configuration and Scripting best practices handy. Get the team to go through it at least twice and once more to make sure everyone knows the game
  2. Quality is everyone's responsibility. Maintain strict traceability on customization and hold team members responsible for deviations. No, I do not mean to thrash them but make them correct mistakes. Pouring over the same code again is punishment enough
  3. Appoint a elite review team consisting of 20% of team members. All "critical" development decisions (do we need a separate view or shall I use applet personalization in the same view?) have to pass through these guys. Also, make them directly responsible for all objects/code reviewed
  4. Organize fun events every month or so during the development period and announce prizes for people who identify deviations from established standards
  5. If your organization has internally developed a configuration review tool, use it frequently during development. If you have no such tool, it is good time to consider developing one
Ofcourse this is not a comprehensive framework, but something I have implemented in a few projects and regret not implementing them in a lot of projects. Do they indeed make sense? What other ways do you follow to sustain quality deliverables? Please comment!