nobalG

[nobalg]

Rant of a Frog

Bruh, Yo Java Guys, why you faking it bruh.

Why you trying to be someone else and failing so hard on it.

May be change is the only way we can be better than yesterday

But your base was so so rigid bruh, its making things difficult for the older community and the newer community is unable to connect with their Grand uncles bruh.

Its obviously clear that at last you accepted the fact that JavaScript's functional way and callbacks were legit from the start. You trying so hard to make Java geeks believe that reactivity is something which will change the world, but I want you to intimate your audience that node js was doing it from so long, concepts like event loop were lying around there from so long. Its just that you were may be late to the party, or have to follow the trend to make the developer's stick to you. May be you want to attract Js developers with your functional styles and lambdas introduction, but this seduction is far from what you planned for.

You only can boost the utilization of multithreaded programs, and utilizing the multicores of a system. But let me tell you this, they will eat you up in this aspect as well.

Why the default was introduced man, in interfaces? Why you had to break your own principles?

Introduce a new concept bruh, but don't alter the story mate, why create such a story, you have created holes in your happy story, while trying to replicate what JS (even though single threaded and have a reputation of having browser language only) have done from long ago.

Functional Programming , ehh, what's wrong with your OOPS man, you used to boost of. Han?

Even though if somebody tries to start from fresh on this path, other then Josh long or Venkat, no legit resources, no community, no answers on Stack overflow for legit questions. Just Blackbox.

So you want to make some threads free ehh, thats why reactivity, ehh,

onErrorContinue ehh, you yourself say the biggest anomaly ehh.

onErrorContinue() design · Issue #2184 · reactor/reactor-core

Not a bug per-se, but a discussion - is onErrorContinue() (and associated onErrorStop()) the right solution, and if not, can the same functi

Please comeback when you are consolidated bruh, We devs dont know how to estimate around you with such open things, We don't know how to handle Exceptions efficiently, Venkat said, if you can't handle or sometimes feels like imperative is better , Go for it.

He said if you want only happy path, then functional will be good.

But ehh, you know right, nothing is happy about programming, there are exceptions , errors , which are meant to be handled.

Bruh, either grow exponentially, or die, as We devs are not able to justify, why you are actually good, but what it will cost to get married with you and your unstability may be.

Angry

[nobalg]

RACE OF THOUGHTS WITH REALITY

Common sense is something we develop, after experiencing different situations. Below are some words, jargons, situations which confused me when I came to the IT world for the first time. They may obviously confuse a newcomer, or may not. But the real problem occurred to me when I already had some meaning assigned to them. Its very difficult to shake of your beliefs, just like people and religion. However some made me laugh, some challenged my psyche, like may be it was better that if they would have been named otherwise. But this may be only my own personal point of view.

I am not going to clarify what the words actually meant in reality, I am just sharing what I felt about them, when I heard them for the first time.

Health Status

The year was 2014, and I joined a company after my college. The manager there was having a team meeting of which I was a part of. I was new and always eager to leave a mark on the org and team, just like some kid who want to outshine the other kids of his class.

In between the meeting the manager gave a statement.

"Why the health status of this agent is so irregular, client is upset about the downtime."

Just this and My brain, has different labelling for this and interpreted this as,

"May be these guys have deployed their personal staff(real human beings-agents) on client locations, who may have been suffering from some bad health and are unable to show up to work. Man, these agents must have been hired for field jobs and thier life seems to be tough."

JavaScript's callback - Unpredictable

The year was 2017-2018, I was having my new interaction with JavaScript, There was a Junior developer with whom I used to work; however he was having more JavaScript experience than me at that point in time.

One day we were writing some code. The code was working sometimes, and sometimes it was not. Actually there was a callback which was relying on some global variables inside its body. Because the other functions were changing the global variables as well, So we were receiving different outputs.

The manager(who had java experience) came and asked the other developer,

"What's taking so long to finish this functionality?" and the developer replied:

"The code is jumping here and there, there must be some problem with the JavaScript"

At that moment, I became afraid as I am just like other humans who are afraid of uncertainty , but I am more afraid of code that cannot be interpreted or can be made sense of.

However, later 'Promises' came into my life after a month and they made me believe that JS world was not actually that bad, and they made me realise that my actual fear was 'Unawareness of basic concepts in JS'

Null values after two hours

This goes back to 2015-2016 again may be. One of the service deployed on dev environment was getting stuck, for at least two hours, then waking up and giving Null Pointers in logs

My team lead(13-14 years experienced) asked us the reason. We told the above scenario and started diagnosing.

The manager came and asked the team lead about the ETA.

He told, "Sir, due to two hours of program being stuck, variables are getting emptied or nulled by something, we are looking into it."

I am not sure whether the manager bought this or not, but it remained a matter of laugh among friends for so many years.

Security

I always understood that securing the application is stopping some hacker to get access to your systems, or identifying some attack on your applications and then recover from the damage etc.

But I realized after going through some spring applications and tutorials, the for a developer it may or may not be same what I used to think.

It was more about Authentication/Authorization/Roles etc.

Hackers were never in the picture(at least while development).

It was really about the Authentication of the user and providing access to the user of the data which he is Authorized for.

It was never like stopping someone working on CMD with green text and typing some commands with pace of light and bringing the whole system down.

It was something else.

UI Development - Pixel By Pixel

For most of the time,(even though I was aware of android at the start of my career ). I always was afraid of Frontend development.

I was unaware of the tools, languages of the UI world.

I used to think, that these screens and the interactions are being written from scratch. May be each pixel is being deigned by some UI developer like a bee creates her home and honey.

Backend Development, everything is Either 'If' or 'else'

Well some developers (old me may be), still believe this. BE is nothing but the ladder of if and else.

Atleast I have heard this from seniors when they assigned tasks to us for a customer change request.

"Oh Man, this is going to be easy, you just have to do this, if this happens otherwise do that, just a game of one or two hours(with unit testing)".

Resource

Use of words like resource for living human beings and then words like allocation surprised me the most. May be only me, but I think human life deserves utmost respect and at least the word 'resource' should not be used to represent an employee.

[nobalg]

console.log(obj) can lie

So if you are doing debugging in browser and doing some printing of objects, list, counters to track some variable or state, then relying on console.log() only may consume some precious hours to diagnose and fix an issue

I will try to keep it short, I was having some Javascript code something like below

1. obj[key] = false;

2. console.log(obj);

3. console.log(key);

4. console.log(obj[key]);

5. and somewhere far away after the execution of above code, obj[key] = true;

The first four lines were kind of close, however fifth was getting executed after certain interval/code lines, but it was sure that it was executing at some point of time.

What I was trying to do was, to track that what value the object was holding at a particular timestamp, and how the value was being mutated with the flow of the program. Obviously the original implementation was a little more complex and the piece of code I have written is just for reference and may include some syntactical anomalies, But you got the point , right? I hope so

To my surprise, console was having output something like below

{

key : true // line2

}

"key" //line3

false //line 4

It was obvious, Javascript and the interpreter cannot be wrong, then why line 2 is showing an object with true in it, but line 4 is false. I thought that may be I am having two different keys somehow in the object, however that was not the case.

I just wanted to track the change in object's value with time. I discussed the problem with my younger brother who happens to be a hardcore frontend guy. He told me that are you mutating the object somewhere down the line. After recalling my memory, I asserted, that "yes, there is something like line number 5 in my code."

He asked me to replace line 2 with below

console.log(obj) ---->--------> console.log(JSON.stringify(obj));

to my surprise, this printed below

{

"key":"false"

}

This time the value was not true.

We searched and found over the internet, that console, actually refers to the reference of object, the unexpanded version of the object in console may display the value of the object at that particular instance when the code was executed, but on expansion inside console, it reflects the latest value of the object, which is kind of lazy loading of object and have no direct mapping with the timestamp(which I think should be the case).

Hope this saves you sometime, incase you encounter some similar stuff.

[nobalg]

Java | static block | beyond

Entry Level definition of static, So static block in java is a construct which gets executed when a class is loaded for the very first time, this block can access the static member variables and static methods of the class.

Any non static member can’t be accessed inside static block and will give error.

The static block is executed even before the main method, however check out the next code pieces attached below for demonstration

However, I was curious that If it is so, then, object creation and normal code, what about these things, I have never seen someone writing any code like this in these static blocks ever, So I was curious and gave it a try. See the output of the code and the comments along side the code carefully.

So as you can see, any type of normal control flow code can be written in static block as well. So I realized that even though the block can have access only to static members, but nowhere in the spec is written that it can’t have its own flow control or object creation mechanism, we can create/throw and handle exceptions in there as well. I just deduced this must be the place where most of the systems would be loading the ready to serve connections/resources, even before any real request is made for some action or data.

[nobalg]

JAVA | Constructor

Constructor cannot be final or static.

Constructor must initialize final variables, if they are not initialized while declaring

Constructor can throw exceptions as well.

If final member instance variable is initialized in try block inside a constructor, compiler will give error that, it may not have been initialized. Even if the initialization ins the only statement in try.

If final member variable is initialized in catch, compiler will giver error that variable may not have been initialized.

Initializing final members inside finally will work, as finally ensures that code will be executed always, unless the finally block is part of some nested try/catch/finally

return statement inside a constructor is valid.

if member variable is static and final, then the initialization of the member variable cannot be done in constructor, The member variable ,must be initialized at the time of declaration.

Constructors can be private.

If no constructors are provided, Java created a Default constructor with empty parameter list.

Once any constructor is defined inside class, Default constructor will never be used.

Constructor overloading is possible.

[nobalg]

JAVA | Quickes

byte   ---->----->> 8bits ---->----->> -128 to 127

short ---->----->>   16bits ---->----->> -32768 to 32767

int ---->----->> 32bits ---->----->> -2147483648 to 2147483647

long ---->----->> 64bits ---->----->> Leave the range, very big number to memorize

char---->----->>   16bits ---->----->> 0 to 65536

Some Gotchas..

Bytes and Shorts are promoted to Ints when used in expressions.

byte b = 2;

b = b*2; //error...Right hand side returns an integer

b = (byte)b*2; //works

int num = 2147483647;

num = num*2; //  num will have value -2

long result = 2147483647*2; //RHS will still return int, so result will have value -2;

long mutiplier = 2;

long result = 2147483647 * multiplier; //RHS have a long operand, hence now multiplication will result in a long

float num = 1.0; //error....every decimal number is double in java by default

However,

float num = (float)1.0 // works, double typecasted to float

float num2 = 1f; //works

float num3 = 1.0f; // works

byte num = 255; //error.... byte ranges from -127 to 128

However, 

byte num = (byte)255; // works, num gets assigned a value of -1

char ch = ‘X’;

ch++;

System.out.println(ch); // will print ‘Y’

Arrays

int array[ ] = new int[12]; 

int[ ] array = new int[12];

int  [ ]array = new int[12];

int array[ ] [ ] = new int [5] [6];

int [ ][ ]array = new int[5] [6];

int array[ ] [ ] = new int [5] [ ]; ///First index is always required

int arrayOne[ ], arrayTwo[ ];

int[ ] arrayOne[ ], arrayTwo[ ];  //declaring two references for ‘two’ 2-dimensional arrays

int [ ]arrayOne,arrayTwo[ ]; //

int [ ]arrayOne, [ ]arrayTwo; //This will give error. dimension in front of the reference can be specified for the first variable only

[nobalg]

Angular Implementing Vertical Steps with animation

I came across a problem statement and I want to share that how I tried to solve it

So look closely, the below component is having a kind of flow to it and it slowly turns from grey to blue down the line.

I went through a couple of approaches and then came to a solution which seems more appropriate while designing this kind of view. I create a repo of the solution, you can check that out from the below link

Github link : https://github.com/nobalggg/vertical-step-progress

Note: If any of the word/term/concept seems foreign to you, you may have to learn/read or google about it, to understand the article.

Most of the time when you touch the smaller areas of a broader field, which seems foreign but in actual they are one step away of opening the doors of possibilities and giving us a broader point of view, provide us with the arsenal and various ways a problem can be solved. Without further ado, I am going to explain below my thought process which I went through while implementation of the above design.

APPROACH 1

I thought of having a table with three columns and then adding adding labels in the left most and right most columns. In the center, I planned to create circles using html and css border-radius.

The next task was to make a line flowing through the circles and also changing the color of the circles along the way. I tried to create a line from a div and giving border to only one side of the div, and somehow placing that border on the center of the table by using absolute positioning. 

The problem was that I was unsure that where to place that line as I wanted the steps to be dynamic and hence the length of the line should be. The view should also respect the container it is placed in and also the size constraints defined by the user of the API. Also to change the color of the circles, a timeout or interval function was planned but all of the approach seems like too much of an effort and seeming like forced solution. 

APPROACH 2 

Now this is the approach which I have implemented, I planned to take a CSS grid which has been recently added to my knowledge arsenal. I divided the Grid into three columns and dynamic number of rows, Along the columns in the center, I placed a canvas ,which lies as a gird item and is spanning across all the rows.

Now rest of the task left was to draw the circles with canvas API at appropriate positions aligning with the step labels which I already added in the left and right of the columns.

For animating the view, I used ‘requestAnimationFrame()’ function. More of the codebase can be checkout out from the GitHub link I posted above. Note that you can increase and decrease the speed of the flow of animation by controlling the below code line in the repository

y= y +3;

if you want to increase the speed of flow, replace 3 with some slightly bigger number and vice versa.

The above example can easily be implemented in plain javascript as well, just checkout the ‘drawCanvas()’ method in the repo. Its written in javascript.

I am thankful that Typescript respects JavaScript code blocks inside its well mannered structure which helped me to save my effort to convert the JS code to angular component with ease.

[nobalg]

TEMPLATE DRIVEN FORMS IN ANGULAR #quicke

Angular reads the html template, and see the form tag as a special case

We want angular to handle the submit of the form which will be triggered by the button of type ‘submit’ present inside the form

We listen to that event by placing (ngSubmit) on the form selector.

A reference with the hash symbol can be placed on the form selector, to create a reference to the form.

That reference should be set to ngForm.

The created reference can be passed for processing inside theTypsecript code, by mentioning it as an argument to the function which is bind to ngSubmit event like :

<form (ngSubmit)=“submitFunction(formRef)” #formRef=“ngForm”>

Above form object can also be referenced inside typescript code using Viewchild.

@ViewChild(’formRef’) formRef:NgForm

The controls present inside the form which angular should consider as part of the form object must have ngModel as a directive.

[ngModel] can be used to assign a default value to the controls inside the form.

[(ngModel)] can be used to create two way binding.

 The properties in the form object will have name as that of the value specified by ‘name’ attribute of html elements.

The properties can be grouped inside a form object using ngModelGroup

The reference to the html elemnents can be created, like below

<input type=“text” name=“username” ngModel #username=“ngModel”>

The reference for the model groups can be created as below

<div ngModelGroup #someModelGroup=“ngModelGroup”> <input type=“text” name=“username”/> <input type=“text” name=“firstName”/> </div>

The form, elements which are have ngModel directive and the groups, all have properties like touched, dirty, valid, invalid, which can help in validating the form

For enabling HTML5 validations which might be disabled by default, place ngNativeValidate on form.

setValue method of ngForm can be user to set the value of complete form and it is directly available on ngForm object.

pathValue can be used to update the partial value of the form and it is available on form property of the ngForm object,(not directly on ngForm)

reset() function on the ngForm can be used to reset the form, which will clear everything.

If reset means setting default value, one can use patch value for the same

[nobalg]

CSS should be respected from the very first day

Different kind of organizations have different people for doing the jobs of web development. Like the traditional approach being followed is  mainly to deliver a web technology enabled product, is to have following categorized teams

FRONTEND: A team of people who will write the code for the user interface of your application.

BACKEND: A team of individuals which will hide all the spaghettis and handle requests which are generated by the beautiful UIs’ ()developed by the frontend team. 

DEVOPS: These people will make the deployment process easy for us, by using technologies like Docker or kubernetes,JENKINS etc. They create CI/CD pipelines to create deployment process smooth.

DATA SCIENTISTS: Whatever data you have, or your application collects, these people will make sense out of that data by using some mathematical algorithms and machine learning tools.

DBA: Database administrators which will handle the allocation of resources/tablespaces/user management/roles over a database

and there are many more categories which I can define and the list will just bore you, however I have listed the main categorization of various teams to deliver a full fledged product for web.

But specifically I want to talk about CSS today, Its not long ago that applications following MVC for example Spring MVC/struts/JSP-servlets etc, mainly handle the Backend/frontend of specific page/feature was done by a single developer.

However due to breakdown of responsibilities, some organizations divide frontend itself in categories.

Framework Developer : Angular/React/Vue

Styling Developer: CSS/SASS/SCSS

User Experience

Wire frame creator: To create a dummy design which will help the investors/Product owners that how the final product will look like.

All these years, I have seen that not all the companies follow these structures. The possible factors which impact these categorizations is funding of the project, number of people in the project with skills, timelines, principles followed by the organization for the delivery of the product.

Because let me tell you this, I have seen(and have done my self) most of these tasks which should be done by the specialists in their domains by myself, But I don’t take it as something wrong, but the inclusion of terms like Full stack and the competition demands from you to be good at multiple skills today.

But I have seen that while interviewing for such positions, people tend to prepare less for CSS and also the organizations hiring for the roles like Full Stack/Frontend technologies ask very less questions from the person in the interview.

I know that stackoverflow/some website/blog will always solve your issue, given that you are not working on some custom created APIs which may have custom exceptions etc specific to the organizations you are working in.

But from my recently gained knowledge about CSS/SCSS I want to tell you that CSS is so powerful technology in itself now, that inclusion of flex and Grid have powered CSS to be taken as A framework in itself now. I was also reading somewhere that animations written in pure CSS are having best performance of all available options and are recommended also. CSS itself have a architecture and project structure for itself for applications. I know JavaScript is magic but sometimes CSS can add more fun to that magic and with very less effort. 

If you are creating class every time in your HTML component for a simple functionality, and surviving on margins/paddings, and you don't know how to center a item on the screen, and your screen breaks on different resolutions; I don’t care if you can handle databases/Rest APIs’/UIs’, you are not full stack developer for me.

[nobalg]

Angular is awesome

Its been a year or so since I started developing Angular applications. Let me tell you this that how relieved I felt after working on angular.

The real reason of this relief is my previous organization’s custom/flawed/framework. Which seemed like it had drawn some boundaries around me. While it was created by the company itself, however the creation of that framework and its incremental growth was a tale of disasters it self. Well just heard the words that the architects failed to create a generic framework that can be just delivered to their domain clients. The generic part was so rigid that any customization which would have been a valid CR was very difficult to implement or sometimes difficult to maintain after implementation.

Well if you are a programmer for at least 5 years or so(even with lesser folks) can agree with me, that global variables/events etc. can be pain at some point of time in applications. While creating modules for the functionalities help us to debug and maintain the code base easily.

Why, because when something wrong goes with these global things, we have to track down and spent a lot of time on the fact that which function have changes the value of the global variable, which component have generated the event which triggered some specific action. As the events may be global and may be originating from number of components.

While with Angular, the first thing the Angular Devs/Team have done is to enforce the Application Developers to follow the best ever Programmatic Development Solution. and i.e creating reusable components.

In simple words, when you want a label in your html page, you will do something like following:

<label>Test</label>

Now you can write your custom events on this, can put validations if it was an input field etc.

Let you have an html page with login form, the login form can be as simple as below

<app-header><app-header>

<login-form></login-form>

<app-footer><app-footer>

While one would argue that these are not html tags I have used above, but this is where the magic is. These are reusable components which can be used any where in the whole application, just like that with their collaborative controls onside them.A component have its very own css respectively.

Their can be components inside components, components can interact with each other and much more. 

However still their are ways to include globalarity(newly created word) or tight coupling among components, but framework wants us not to got for it. 

[nobalg]

Why learn Algorithms?

Algorithms, Do I need to cram them?

Do I have to learn the steps of already existing algorithms?

Why do I have to learn them, if everybody can find them easily by googling them.

Do I have to learn them, just to crack some interview?

Well, a lot of questions; I rarely got any chance to write them down In my professional life till now, because the languages like Java and others are so enrich in APIs’ and providing the implementation at there core for the algorithms, and the Data Structures. I just have to import particular package and yes, I am done.

However the world of algorithms and Data structures holds an important role in the tech world. Data is getting bigger and bigger. Access to the data needs to be fast. Hence knowledge of these algorithms is a must parameter for techie to design new solutions/algorithms for the existing/ upcoming tech world problems.

Learning algorithms/Data structures is the process of setting and training your mind to pick up the appropriate set of decisions towards solving a problem.

I am going to discuss a very simple problem which will depict the same.

Consider the following question,

From a given array of integers, find the ‘k’ max numbers. Let the array be [17,25,85,46,32,45]. Then for k = 2, output will be 85,46. You are not allowed to use Collections API or any other utility methods.

How will you approach to this problem?

A naive approach will be to loop through the array, find the max, save it somewhere else. Then find the next max. Well this is the first thought of every programmer not having knowledge of Algorithmic approach. Well it may yield the result for you, but its not efficient.

Well what we have to do is, just to sort the array using any algorithm, and get the last ‘k’ elements from the array. Now comes the question which sorting algorithm we must go for? Let that I have merge sort and bubble sort to choose from. Which one will you choose to solve this problem.

What is Merge Sort? Check here

For Bubble Sort, Go through this

What I learned about the sorts is that Merge Sort is one efficient algorithm, having time complexity as O(n log(n)), and Bubble Sort is the algorithm having the complexity O(n2). Let me tell you that these are the time complexities the algorithm will take on the basis of the input size. As log N is less than N, so the time complexity in case of merge sort is less.

So clearly merge sort should be my pick, but No, bubble sort will be performing better in this case for smaller ‘k’ values in the above problem. Because , we are not asked to sort the whole array; but just to pick the max ‘k’ values.  In bubble sort, the core for loop will run up to only ‘k’ times and will give us the final result more quickly. And there will be very less trade off for memory also, as merge sort uses more memory for storing the temporary values while performing split and merge operations.

What I want to depict is that that  lot of years of practice, patterns, and dedication will tell you that how should you approach a technical challenge, Which and what combinations of algorithms, one should use. May  be you will end up creating better algorithms from existing one. Or Create a new one for some different set of problems which exists but have not  been discovered yet. 

[nobalg]

ANGULAR ( App in a browser?)

Let me get this straight that angular apps, are not  normal web apps and are single page applications.  What I define An Angular app is as a, loosely coupled set of reusable components brought together to provide user with a fast and easy access to various services over the web, without reloading a page, and just some REST/ Ajax calls to get the data chunks.

If you are good at angular, you probably know better than above and below this paragraph, however follow for more simplification and my opinion on angular as I am half way through it.

What are apps? Why are android apps are ‘apps’, not web apps, and angular apps are ‘apps’ and not web apps? Because in the world of software development,  Apps are referred to as piece of software, whose most of the code and logic is on the machine where the client is interacting with user interface.

If you ever tried to write code for angular(as of angular 6), you will write your code in pretty cool language developed by Microsoft known as ‘TypeScript’. and  in case you are using editor like VScode, let me tell you that it is like sliding on a boat in a river full of butter. May be I am getting too emotional, but believe me, I loved this combination.

TYPESCRIPT

Typescript is super set of Javascript. I was not happy(will not say hate) with the absence of classes and objects in Javascript (yea objects like java). But Typescript made me feel at home. A constructor in type script class is  a function called as constructor() And Oh boy look at the below syntax,

and what if I tell you, below will create properties like age and address in the constructor itself, and will behave similar to the above code.

and if you create a property in a constructor and used some existing class of angular/or your own, then with the magic of ‘Dependeny Injection’, that object will be available in your code, like the following

SINGLE PAGE APPLICATION

Does this mean, URL will not change through out the app life cycle?This is what I was most curious about. Just like those shopping websites, where you scroll and scroll and they keep loading data in batches.

Na, it can and it can not. You can handle this logic in your code with the RouterModule present in angular, which will intercept form submissions and links which you will provide in the angular app itself according to the logic written by you. We don’t navigate through ‘href’ links, but provide our own configurations in HTML tags to handle navigation and can do the same programmatically. 

FORMS

Forms can be created by two ways in angular, and no server submission is there in both cases.

Template Driven Approach

Reactive Form Design

Let me explain this to you in simple words.

Template driven is more configuration in HTML code , less code logic

Reactive design is more of code, but less configuration in HTML.

Many more is there, but I generally come up with the Trailers of the whole movie which I have already enjoyed and motivate others to dive. Will share my thoughts on angular in future too. Peace.

[nobalg]

TOM to BOSS #JAX-RS

I was comfortable with Tomcat because it was so simple and understandable and configurations were so easy, then moved on to JBoss, It was having some major features; along with those, come complex configurations and customizations, but if you understand them well then you will know that why JBoss is called as  Application Server and Tomcat is in category of WebServer

I tried to deploy one of my functional Jersey Rest web service implementation on JBoss, but unfortunately the deployment failed.

 I was surprised at first, because the services were working well on Tomcat. So was wandering that what went wrong with the JBoss. I searched and went through some articles over the internet and gathered some sequential steps to get my work done. While following the solutions , i came across few new errors/exceptions, so here goes the list and their solutions:

Only One JAX-RS application allowed

When my war file failed to deploy in JBoss, I ran an enquiry on logs and came across this line in the Server log, “Only One JAX-RS application allowed”. Its ambiguous message. It seems like it depicts that only one war file with JAX-RS application can be deployed, but this can become a major flaw of JBoss if it is unable to run multiple JAX-RS applications.

But the real issue was something else. The issue was the REST-EASY implementation already provided by JBoss. I was creating the Rest Services using jersey and hence deploying them on JBoss, but JBoss was already shipped with REST-EASY packages. So I need to tell JBoss by some way that use jersey and ignore the REST-EASY implementation. So for doing that, I added  following lines in my web.xml file .

<context-param>    <param-name>resteasy.scan</param-name>    <param-value>false</param-value> </context-param> <context-param>    <param-name>resteasy.scan.providers</param-name>    <param-value>false</param-value> </context-param> <context-param>    <param-name>resteasy.scan.resources</param-name>    <param-value>false</param-value> </context-param>

By adding so, JBoss was atleast understanding that which implementation is needed now, but a new issue was awaiting.

java : Unsupported major.minor version 52.0

Now what?, I said to myself. One gone another came. But in an instant I asked myself, is there any difference between the JRE of client machine and the machine on which the webservices were compiled. And there was the resolution of this issue as well. I was compiling the services in JDK 1.8 and deploying them on a system having java 1.6. I recompiled my services and deployed them, and the error was gone. But this was not the end yet.

Package Scanning is not possible

As I progressed,  I came across another issue in server log. There was no way the server is going to package scan my Java packages to know that which service classes I have implemented. In other words,for everything to work as expected We have to annotate our classes with some special annotations to let servlet container know that which are web service implementation classes. I had already done that. But this new issue came up along with a suggestion in the logs. The suggestion was to manually implement an application class by myself and register the Resource classes in it rather then depending on annotations. So I added something like following into my existing package.

and the error was gone.

Issue with the Filters

As explained in the above issue that I had to register my rest resources(classes) in application class itself, In similar way I registered my filters there and added them in the existing set. However the filters were not getting executed even after doing so. Hence I removed the filter from application class and mentioned them in web.xml of my project. Magically they started executing.

No more..........

After getting these issues and resolving them one by one,my half a day was spent but I was able to get them working and deliver the build on time. The more exciting thing about this chain of issues was that; this was my first rest deployment project and was an immediate requirement by the client. So, I was happy that I managed to get that in time and so I am sharing the my experience, so that if any of you may come across these will be able to solve them.

Thanks for reading...

[nobalg]

You can but shouldn’t #Java #static

In any country, if you are driving a bike or a car, then there are rules which you must follow. These rules are nothing but conventions which ensures each traveler’s safety. If everybody follows them , no accidents happen. A slight ignorance can lead to disastrous accidents. You can take your car over a footpath, can jump traffic lights but as the title says, you should not. At times you may get away with these stuff but still by not following the conventions and rules you are putting yourself as well as others in danger.

I write a blog post every time I find something interesting or worth sharing. In case of static methods in java, there is stuff which is allowed and no one is gonna stop you from doing , but by convention we must avoid those scenarios. Before introducing the ambiguous scenario 

lets understand static first.

Understanding static in Java quickly

static members belong to the class instead of a specific instance.

It means that only one instance of a static field exists, even if you create a million instances of the class or you don't create any. It will be shared by all instances.

Since static methods also do not belong to a specific instance, they can't refer to instance members (how would you know which instance Hello class you want to refer to?). static members can only refer to static members. Instance members can, of course access static members.

static members can access instance members through an object reference.

SCENARIO

Lets say you have a class ,

Now you know that static methods don’t need an instance of a class to make a call and can be called directly from class name as follows.

and also, static methods of the class can be accessed by the reference of the class too.So following code is valid too.

Whats strange is the following piece of code, it will compile as well as execute successfully, The sysout of the method body will be printed in the console.

How come? The object reference was null and the method was called on a null reference . It should throw NullPointerException. But the program executed successfully.Why?

Now, Repeat with me, 

“When accessing a static member through an object reference expression, only the declared type of the reference matters. It doesn't matter if the reference is actually null, since no instance is required“ .

If you would have called a non-static method on a null object reference, you would be getting the Null Pointer Exception, but not in this case. It was because your method was static.

You shouldn’t ever try to call static methods on object references rather they should be called on the class names only, as they belong to the class, not to the instances of the class.

Why should you avoid ?

Because  it gives a very misleading appearance on what it's actually doing. For instance think of it as you are reading someone’s code to make a fix or feature addition. You came across something like this.

while adding your feature, you will be thinking that the ‘object’ reference is not null, and hence you can add other method calls below this piece of code. Hence you added a call to your own newly created non-static method below this code like following

what if the someMethod was a static method and object reference is null. Your code will fail in that case.

Hope you understood the topic. Thanks for your time.

[nobalg]

FACADE PATTERN #java

Person 1 : Hey hows you mate?

Person 2 : I am more than fine buddy.

Person 1 : Thats’ cool, Hows your life? Everything is good?

Person 2 : Yea mate ! sort of the same everyday.

Person 1 : Oh well! lets crisp some things up by learning something new, 

can you write a Java class for me that will depict your daily routine of your day in methods in a package named facade.

Person 2 : Haha..not again, What are you up to?

Person 1 : Do it man..... will tell you soon.

Person 2 : Ok, so here it is...

Person 1 : Well you have done it right, but Man I don’t want to know your daily routine but still want you to live for me, spend your day for me. So without further ado I am going to create another class from your class.

and when ever I will want you to spend your whole day, I will not have to know that How you are going to spent it, but still your day will be spent in the same manner you want. So following piece of code will hide the complex code for me and will still work like charm.

Person 1 : Just above code and I can make you spent your whole day.This way of hiding complexities in your code and wrapping them under some other class is a special design pattern and is called as FACADE DESIGN PATTERN.

There are different ways in which this design pattern can be implemented , however the above given example is the best and easiest. This pattern is mainly used to simplify the existing classes and methods and bundling up the behaviours and workflows of consecutive methods in one place or more than that. 

In the above example, Person 1, don’t want to know about the schedule of the P2 every time and have written a class which can hide the complexities for P1 and he still can code the functionality with ease. Facade pattern is one of the simplest patterns, and I  hope you understood this well. Thanks for your time.

[nobalg]

‘REST’ in peace... #JAX-RS

CAUTION : The article solely consists of the developer talking to himself about the JAX-RS and its features without much details,This is occasional flow of emotions. So if you want to learn JAX-RS, try reading a book or Tutorial. 

RESTfu​​l Jav​a​ wit​h ​JAX​-​​RS 2.​0​ (Second Edition) by Bill Burke is the way to go

When I hear the words like, scale, abstraction, encapsulation, polymorphic, logic, magic, communication, interaction, doubt, ambiguity, efficiency, systematic, and many more similar words, I just fell for the topic regarding which these are being used for, I don’t know why, I guess every normal human will.When I hear that “The particular ‘subject’ you are looking at, serving you with services without letting you know that how it is doing” are one of the few quotes that shake my mind and tell me to dive in the concept.

So if you will ever make up your mind to learn about the JAX-RS  , believe me whatever book or tutorial you will pick, the first chapter will have that bold dark line in it. Annotations , customisation and each feature you will  go through while learning will be fun stuff and will never bore you,(unless you are not practising the small code examples along the way).  

JAX-RS utilises the already existing HTTP specifications, which are the backbone of all web interactions happening over internet all over the world to create and share Resources on servers.So you know its easy for someone who wants to learn how to ride a a bike who already knows how to ride a simple scooter(atleast easier than learning some out of the world new vehicle).

You just need to stick to the basics, the accelerator is still there,the  brakes are there but the clutch is added, You can learn it with practise right. 

I liked the idea when developers of the Android framework picked up JAVA as there primary language, (although under the hood there is no JVM , or byte code is involved. ) as JAVA community is one of the largest developer community across the globe. Developers learned the skills quickly and now Android Developer community has grown in number  considerably.

Back to the Road, REST is so easy in comparison to SOAP, and implementations like Jersey, RestEasy and RestLet are the way to go with it. Application endpoints are  easy to develop and manage.  

Whenever I see some Job posting in Android or any other  profile, JAX-RS is one of the requirements.  Web is prevalent, and hence web services are. JAX-RS also provides cross points called as filters and Interceptors, that will make life easy for you when you think of reusing some code across your applications.

BUT...

The one and only pain I found in the API implementation, was the long running arguments in the methods.For example see the syntax of following abstract method of MessageBodyWriter

public void writeTo(TableData data, Class<?> type, Type genericType, Annotation[] annotations,MediaType mediaType, MultivaluedMap<String, Object> headers, OutputStream out)

Cmmon Guys, you know Robert Martin is gona hate you for this, he clearly mentioned in his book that avoid these long running argument scenarios, However, Lets just compromise for now.

I hope you will also love to learn it, Do some code practices along the way and you will master it easily. I am polishing myself too.

Thanks for Reading

[nobalg]

Yea, I am in the market... #AndroidApp #TossIT

Its not long ago (seems like) when I used to develop small pieces of code for myself to play with and make small Java, Android applications. They were useful in someway or another but not 100% complete. Irrespective of the standard application development procedures, which we follow at my organisation where I work, I knew that these small pieces will be on my hard drive and will remain there for me as learned educational experiences. Hence my efforts were limited but I was adding my own crunches and crisps to my own little code snippets.

However last night I had a long chat with one of my Great friend(not best) about the Android applications published in Android Market by the company where I work. We were discussing the Goods and Bads’ , the use of intranet android applications uploaded by my company and etc. In between our talk full of buts and ifs, he suggested that why not having a developer account of your own on Google Play and upload whatever small piece of application I use to create. It stuck me that I was developing Android applications for production since last two years, but haven’t uploaded anything of my own. So, why not buying an account and start uploading some simple , old applications which i used for my learning.

Hence here we are

https://play.google.com/store/apps/details?id=realmknight.toss_it&hl=en

The above link will redirect you to google play to my application which I developed, I guess last year. Its a one page application of a flipping coin. i I named it ‘TOSS IT’.You can use this free application as a replacement for the real world coins you use for ‘Toss’ before any sports to make certain decisions .Track your own ‘Luck Factor’ if you are alone.

How to use It?

Simple, Shake your phone up, down, right , left, flip coin make your call, HEADS OR TAILS..............and there you go.

While developing and uploading it I learned a few more things in practise, some of them was new to me.But all of development was a great experience .

I am planning of uploading more of my small applications by enhancing them soon. Will try to create entertaining as well as useful stuff. Try it and do comment your opinions, Peace..