I’d used the online funds transfer facilities of my bank many times before using the NEFT method. This time though, I stumbled upon the RTGS account of DELL. I was curious since I’d been using NEFT all along.

So I did some research and here’s what I found:

The basic requirement for transferring money electronically on your own is to have a NetBanking Account from your Bank. Almost all private banks offer the service of a NetBanking account to their customers freely. The only thing you need to do is to fill in the form and mark all the services you need to avail on that form.

The first step to take before making an online transaction using this account is to add a payee to your list of parties with whom you would like to carry out these transactions. This can be done very easily by providing the Bank and Account Details of the concerned party. Details provided by you are verified and the party is a added to your list of payees in a maximum time of 24-hrs after which you can perform your transactions.

EFT or electronic funds transfer to accounts within the same Bank is done very easily and it takes place almost instantly. But transferring money to accounts outside your own bank takes more time. Funds transfer to other banks can be done in two different ways that serve different kinds of needs and functions. They are:

• NEFT – National Electronic Funds Transfer
• RTGS – Real-Time Gross Settlement.

NEFT is suitable for customers who want to transfer relatively smaller amounts. This kind of transaction may take something between a day or 3 days to reflect into the payee’s account. This is because NEFT operates on a Deferred Net Settlement (DNS) basis which settles transactions in batches. In DNS, the settlement takes place at a particular point of time. All transactions are held up till that time. For example, NEFT settlement takes place 6 times a day on weekdays (9.00 am, 11.00 am, 12.00 noon. 13.00 hours, 15.00 hours and 17.00 hours) and 3 times during Saturdays (9.00 am, 11.00 am and 12.00 noon). Any transaction initiated after a designated settlement time would have to wait till the next designated settlement time. NEFT transactions attract a maximum charge of Rs 5/- for a transaction below Rs. 1 lakh and of Rs 25/- for transactions above Rs. 1 lakh.

RTGS system is primarily for large value transactions. The minimum amount to be remitted through RTGS is Rs.1 lakh. There is no upper ceiling for RTGS transactions. Also, under normal circumstances the beneficiary branches are expected to receive the funds in real time as soon as funds are transferred by the remitting bank. The beneficiary bank has to credit the beneficiary’s account within two hours of receiving the funds transfer message.The RTGS service window for customer’s transactions is available from 9.00 hours to 16.30 hours on week days and from 9.00 hours to 12.30 noon on Saturdays for settlement at the RBI end. However, the timings that the banks follow may vary depending on the customer timings of the bank branches.

RTGS transactions attract a charge of Rs 25/- for a transaction between Rs. 1 lakh to Rs. 5 lakh and of Rs 50/- for transactions of Rs. 5 lakh and above. Also note that RTGS facility is not available with all the banks. As on December 31, 2008 more than 52,000 bank branches are RTGS enabled. The list of such branches is available on RBI website.

The debit of the transferred amount from your account happens instantaneously, the credit to the beneficiary account happens in some time, depending on the kind of transaction. In case, the transfer is unsuccessful for various kinds of reasons the transfer amount is credited back into your account along with any fee charged for that transaction.

We first looked at RadiantCMS, with the intention of building our web application on top of it. But then decided to instead build our web app from scratch and then make parts of it editable using Comatose [which is a "MicroCMS"]. This way, we have more flexibility to add features to the web app using the Rails API and available plugins, rather than being constrained by RadiantCMS system.

Things we liked about Comatose:

• Its implemented as a Rails plugin – which makes it a breeze to install
• Implementing Single Sign On is also simple. Because it works well with the  restful_authentication plugin.
• Comatose allows rendering of pages inline from your application view – just like rendering a partial in a Rails app. To include an “editable” tool tip within a form, something as simple as the following is enough:

  • <%= render :comatose=>'about' %>

There’s of course a lot more you can do with this plugin. Check it out.

Now for the bad part

• The main issue is font inconsistency. There are some options for setting the font size, but they don’t work as expected – because the CSS styles are inherited from our app.
• In production mode, there is a noticeable delay between making changes in the CMS and the updated content showing up on the main site. Not sure why this is the case.

• Frequent Validation – Always validate with actual customers. Talk directly to them, take their feedback, implement it and soon release it. This cycle has to continue. If we don’t do this, there is no meaning in building any product with BDD/TDD, Automated tests etc.
• Limiting WIP [Work-In-Progress] – Never leave too many things untested or unreleased. Get your QA team to test as soon as the dev team is done. Release to production as soon as QA is done. When we make frequent releases, you can also get frequent validation.
• Checking vs Testing – If you’re not sure what the difference is, there’s an article on InfoQ that explains this well. The bottom line is, use automated tests for mundane tasks ["checking"] and use manual testing for exploratory testing.

There was an interesting session by Siddhartha Govindaraj, who is the founder of Silver Stripe Software. He talked about how they evolved from ad-hoc to Agile and from Agile to Kanban.  The points I felt interesting and which are worth trying out are:

1. No iterations/sprint - Always take top items from backlog – What usually happens during an iteration is, either you might finish all the items in an iteration and as there is still more time left you might have to pull some item from the backlog. The other case can be, you are not fully done with some tasks you might have to push some items to next sprint. Instead of this, always take items from backlog. The developers are supposed to work on the top items in the backlog and QA will be testing as soon as the development is done. QA need not have to wait for a certain period to test.
2. Limiting WIP (Work In Progress) - Don’t pile up stuff, never leave anything untested/unreleased. If WIP goes beyond a specific number, then change the plan like stop development make devs to test. And don’t keep tested stuff unreleased. Alway keep a maximum number of items that needs to be part of released. So even if the release is planned weekly, there can be multiple releases during the week if a lot of features are implemented during a certain week. Another advantage with this approach is, when dev team gets a chance to sit with QA, they also learn about exploratory testing.
3. Single backlog for multiple projects - This is an interesting point. He had multiple projects say A & B, both in maintenance stage. He has a team of 5. So rather than splitting the team into 2 across these 2 projects, they kept a single backlog of both projects. They prioritize the backlog and take items from that. This way both projects move in parallel more smoothly.
4. Checking vs testing - Checking should always be automated while  Manual testing effort should be reserved for exploratory testing only.

He also suggests avoiding a daily Scrum meeting. The point he had was, the team should be interacting so closely throughout the day, which avoids the need for a stand-up . But if there is an issue that needs to be discussed or if there is a requirement for having a discussion within teams, then such a stand-up is called for. I heard the same being discussed by many people – that Scrum is not really mandatory but am not sure about the same as of now. At Multunus, we still get quite a lot of value out of the stand- ups, because we do it a little differently.

It was a pretty simple presentation in which he was talking about how their company has evolved. As it was about his own experience, he had concrete examples for validating his points. There were questions from the audience like, how velocity is calculated if there were no sprints, how the team would get adjusted to context switching in case of single backlog etc. His answer was that, they are a small team – so these were mostly non-issues for them. And about velocity, productivity etc, those really will come into picture only if there is not enough trust in the team. If the team trusts the management and the management also trusts the team, then these productivity/burn down charts etc are meaningless. That point really made a lot of sense to me. If the team always perform/deliver well and if the management has trust on the team, why should they look at the burn down charts etc to see the team’s performance?

Do you like this list? Or not? Have suggestions for more? Go ahead, talk to us in the comments below. We’d LOVE to hear from you.

1. Export the certificate from IE by following the instructions given here: http://www.tech-pro.net/export-to-pfx.html
2. The certificate with Private key will be exported as PFX format in the above step – but this cannot be used by the jarsigner. So you need to convert it into “p12 format” which the jarsigner can understand. For that you need to use Mozilla. First import the certificate  saved in step 1 into Mozilla as follows:
   1. From the “Edit” menu select “Preferences” and open the “Privacy & Security” category and click on the “Certificates” item.
   2. In the “Manage Certificates” section, click on the “Manage Certificates” button.
   3. In the “Certificate Manager” window, the “Your Certificates” tab should automatically open. (If not, select it.)
   4. Click the “Impor”t button at the bottom of the window.
   5. It prompts you to select an existing file; select your certificate file from the location where you saved it in when you exported it.
   6. It prompts you to provide the “Master Password”; enter it, if you have set one.
   7. It prompts you to provide the password used to encrypt the certificate backup; enter it.
   8. It should say “Successfully restored your certificate(s) and private key(s).” Click OK.
3. Then export the certificate as p12 format:
   1. From the “Edit” menu select “Preferences” and open the “Privacy & Security” category and click on the “Certificates” item
   2. In the “Manage Certificates” section, click on the “Manage Certificates” button.
   3. In the “Certificate Manager” window, the “Your Certificates” tab should automatically open. (If not, select it.)
   4. To export your “DOEGrids” or “KCA Personal Certificate”, click on it to select it, and click the “Backup” button at the bottom of the window.
   5. You’ll be prompted to specify a filename and location for the “PKCS#12-format” certificate file (file extension will be “.p12″ in UNIX/Linux,  ”.pfx” in Windows).  Provide them and click OK.
   6. A dialog box requesting the Master Password may appear (the password and certificate database). If you have set a Master Password, provide it. If not, you can make one up and provide it (optional). Remember this password!
   7. You’ll be prompted to make up and (twice) enter a second password. This one is for restoring this particular backup of this certificate. Remember this password!
   8. Once the system says it’s successfully backed up your certificate and private key, click OK.
4. We also need to know the alias of the “.p12″ file so run:

    keytool -list -storetype pkcs12 -keystore <path to the cert file>

5. Then you will see output like this:

Keystore type: pkcs12 Keystore provider: SunJSSE

Your keystore contains 1 entry

xxxxxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxx, Apr 26, 2006, keyEntry,

Certificate fingerprint (MD5):hh:hh:hh:hh:hh:hh:hh:hh:hh:hh:hh:hh:hh:hh:hh:hh

The xxxx-xxx… number is the alias for the key

6. Change your ant script as follows to sign the certificate:

<target name="signjar" depends="jar">
 <signjar jar="yourJar.jar" storetype="pkcs12"
   keystore="yourkey.p12" alias="Your Alias"
   storepass="your password"/>
</target>

JQT is a family of JQuery plugins comprising of the most common widgets used in “modern” sites.  We’ve used quite a few of the widgets on our sites in the last 3-4 projects and it  has certainly helped us improve the look and feel of these sites – with much less effort that what would have been required otherwise.

Installing and using JQT on your site is also super simple.  Download a JS file of ~6kb and include in your web page.  The documentation and demos provided for each widget is excellent – so you should be on your way to actually getting stuff done very quickly after the download.

Bottom line – you’ve got pretty much no excuse, to not check it out .

1. We want to reuse application code built for a .NET application
2. We want to integrate the DLLs with a Java based sever technology
3. Should be simple enough to maintain the code by .NET professionals

There are couple of Java-COM bridging solutions I came across.

• Java Native Interface (JNI)
• JACOB
• Scriptom (Groovy – COM Scripting)
• Java Native Access (JNA)
• JNBridge

In this post, we’ll see how Scriptom can be used to access DLLs.

Scriptom is a module in Groovy (A Java VM based scripting language) which wraps the JACOB classes to make it more usable. Scriptom’s syntax looks like VB.NET code.

Lets jump in to the sample application.

Assume that we have a VB class named TimerState in the TestDLL VB.NET project, exposes the COM, signed and generates the DLLs.

Following is the TimerState VB class. In this class, we have an event UpdateTime which raises an event when to notify updated time.

 <Microsoft.VisualBasic.ComClass()> Public Class TimerState

    Public Event UpdateTime(ByVal Jump As Double)

    Public Sub TimerTask(ByVal Duration As Double)
        Dim Start As Double
        Dim Second As Double
        Dim SoFar As Double

        Start = Timer
        SoFar = Start
        Do While Timer < Start + Duration
            If Timer - SoFar >= 0.1 Then
                SoFar = SoFar + 0.1
                RaiseEvent UpdateTime(Timer - Start)
            End If
        Loop
    End Sub
End Class

We want to use this logic in the Scriptom code and do something whenever a UpdateTime event is raised. Here are the steps:

1. Register the DLLs (this is required step for the Java Virtual Machine to get meta data about the DLLs)
2. Download Scriptom
3. Copy the jacob-*.dll to any system PATH
4. Have all the JARS from Scriptom jar folder available in the CLASSPATH

Following is the sample code for invoking

import org.codehaus.groovy.scriptom.ActiveXObject // Imports the ActiveX object class which holds the reference to the class in DLL

def timer = new ActiveXObject("TestDLL.TimerState") // Looks for the meta data in the registry with "TestDLL.TimerState" name and creates a reference to that class

def time = 1.00

timer.events.UpdateTime = { variants -> // event registration for UpdateTime event in the VB class. Variants holds the data passed from the event. In this case, it is Jump value.
    println "UpdateTime: " + variants[0] // We're just printing the upated time
}

timer.TimerTask(time) // starting the timer task in TimerState class
while(true) { // keeping itself alive to respond when timer events are raised
}

In the next post, we’ll discuss about JNBridge, a comercial solution.

We’ll be blogging about some of the above topics in more detail in the coming weeks. Do drop a line in the comments section below, if you’d like to learn more about anything specifically.