Very rarely, you will want to delete all friends from Facebook and start a “fresh” facebook account.

Now, it’s hard to do this manually. Thanks to phpunit + selenium + a few lines of code, this can be done now.

The following script removes all your friends from facebook Replace and with your username and password respectively.


<?php
require_once 'PHPUnit/Extensions/SeleniumTestCase.php';
class Example extends PHPUnit_Extensions_SeleniumTestCase
{
  protected function setUp()
  {
    $this->setBrowser("*chrome");
    $this->setBrowserUrl("http://www.facebook.com/");
  }
  public function testMyTestCase()
  {
    $this->open("/");
    $this->type("email", "<username>");
    $this->type("pass", "<password>");
    $this->click("//input[@value='Login']");
    $this->waitForPageToLoad("90000");
    for($i=0; $i<1000; $i++) {
        $this->click("link=Profile");
        $this->waitForPageToLoad("90000");
        $this->click("//a[@href='http://www.facebook.com/sp2hari?sk=friends&v=friends']");
        $this->waitForPageToLoad("90000");
        $this->click("//div[@class='fbProfileLargePortrait']/a/");
        $this->waitForPageToLoad("90000");
        $this->click("link=Unfriend");
        sleep(5);
        $this->click("remove-friend");
        $this->waitForPageToLoad("90000");
    }
  }
}
?>


Note:
1. The script will stop in between with timeout error. You probably should run this script in an infinite loop.
2. Once you’ve reached 0 friends, facebook blocks you from adding friends in bulk for 2 days. So, you probably have to wait till facebook trusts you as human and not a bot.

Happy hacking

VIM is an awesome editor, but comes with a big big learning curve. These are the the top 3 Vim tutorials from where I learnt to use Vim.

1. http://www.vi-improved.org/tutorial.php

2. http://vim.runpaint.org/toc/

3. http://www.swaroopch.com/notes/Vim_en:Table_of_Contents

If you want to learn more, checkout http://thomer.com/vi/vi.html#manuals

Git reflog saves you from all screw-ups you do while using git.

Here is a sample output of git reflog.

hari@hari-desktop:/var/www/recruit$ git reflog
e458f54 HEAD@{0}: checkout: moving from test1 to e458f54fe99202c82c2690be4fa2fe2d7aa7be32^0
a3281d1 HEAD@{1}: checkout: moving from master to test1
f265802 HEAD@{2}: checkout: moving from test1 to master
ef08492 HEAD@{3}: ORIG_HEAD: updating HEAD
e458f54 HEAD@{4}: commit: Implement phpunit testcase.
1ecf2c3 HEAD@{5}: commit: Bug#1 : MCQ Choices displayed in questions.
f265802 HEAD@{6}: checkout: moving from test1 to f265802d1d65a9ac92bf9f6414d8f468444cce66^0
ec280c6 HEAD@{7}: checkout: moving from release to test1
f265802 HEAD@{8}: merge master: Fast-forward
ccec6c4 HEAD@{9}: checkout: moving from master to release


All your commits, checkouts, pull, merges are logged in here. To, reset the git to any point in there, we use the following command.

git reset --hard <commit-id>

But if you push this to the git, you will find that the commit is rejected now. To forcefully push your new HEAD to server, we use the following command.

git push origin +<branch_name>

+ before the branch name forcefully pushes the branch (even if your current branch is behind the origin).
NOTE: Please be careful when you use the command above. You might lose all the changes made by others (temporarily) if you use it. Of course, using reflog + reset, you can get back the changes.

Want to check if all changes in your new branch is merged to main branch before you delete the new branch.

git log feature1 ^master

The above command displays all commits which are present in feature1 and not present in master.

Note 1: You can also try “git branch -d feature1″. If the new branch is not totally merged, git won’t delete the new branch.

Note 2: You can also try “git branch –contains feature1″, but I’m not sure how that works. I guess I’m happy with git log for now.

To read more about this, checkout http://stackoverflow.com/questions/1710894/using-git-show-all-commits-that-are-in-one-branch-but-not-the-others

Hiring good candidates is a problem for every single company out there.. no matter small or big. We at interviewstreet attack the tip of the iceberg by filtering good programmers from the rest. People say, it’s a huge timesaver .

But enough about interviewstreet, the first problem companies should tackle is to make sure rockstar hackers apply for the job. If they don’t apply to your company, then hiring them is really impossible.

So, how do we make sure awesome rockstar programmers apply to your company?

Have an awesome “Careers” page.

Please repeat the above statement 10 times. Bad careers page doesn’t attract motivators or problem solvers. Let’s see a few nice careers page I found in the web.

1. Google [ http://www.google.com/intl/en/jobs/profiles/whygoogle.html ]

I just loved this page. They explained why search is hard in such a way that even my sister could understand that they are doing awesome work there. Next, they have explained what they’ve done till now. The sections below that excite the programmers who are planning to apply. An awesome careers page, which a programmer and a school kid can understand and would want to apply. Certainly, their whole Jobs section (http://www.google.com/intl/en/jobs/index.html) is worth looking it.

2. Amazon [ Careers-University-Recruiting ]

The tag line is the best. Work hard. Have fun. Make history.. It’s hard to resist applying for amazon after checking their Why Join Amazon? page.

3. Dropbox [ http://www.dropbox.com/jobs ]
If you think you need to be an MNC to add the perks and other details for an awesome careers page, then change your mind now. The benefits and perks speak out their startupish culture. Whiskey Fridays, Gaming, and music in office. Why wouldn’t someone work in a place, where you solve the biggest problems in between whiskey, music and gaming?

4. Redbus [ http://www.redbus.in/careers.aspx ]
Neat description about the company.

So, what’s the ingredients for an awesome careers page?
According to Sameer from TheMorpheus,

1. The job description should be exciting and full of energy. If you aren’t excited about startup, don’t expect others to be.
2. It should be clear that the company would be fun to work for.
3. The job description should be very focused on appealing to folks who are usually hackers. Make sure you talk a lot about technologies you are using and you are planning to work on.
4. Be very upfront about things. If you want folks who want to do good work and not just looking for money, if you expect people to work on weekdays and almost 24 hours a day, mention it directly.
5. Most of job descriptions are plain boring. Make sure the job description is honest, personal and connected to person reading it.
6. Talking about traction and founder background you have is good but not special because everyone writes that now.

I would like to add a few more points to the above list.
7. Don’t ask questions like are you a super man, can you do super duper jobs etc. I personally feel it is just wasting space in careers page. For example the careers page in IBM [ http://www-07.ibm.com/in/careers/index.html ] , questions like “Are you a problem solver? Are you a difference maker? Are you ready to make our planet work smarter?”. Even though these questions talk to you directly, the value it adds is not that much. No one is going to answer these questions with a “No”.
8. Don’t make too long or split it to many pages, unless you really can’t avoid it. Remember KISS.

Salary

Salary plays a major role in attracting candidates to your company. But you need not compete with MNCs if you can’t afford it. Flipkart’s career page, explains this clearly.

We are small, but we plan to offer all that big companies do – Competitive salaries and perks. We want to get money out of your head so that you can concentrate on your free will and creativity. We are working on setting up other benefits like Provident Fund, Medical Insurance, Gratuity, Super annuation etc. We distribute goodies every now and then.

So, what’s your careers page like? Feel free to add your careers page link in the comments section.

HTML5 seems to be the buzz word now. People around me are excited about HTML5 or already working on something with HTML5.

Curious to know what the hell is this HTML5, I started reading something from https://developer.mozilla.org/en/HTML/HTML5. Well, truth be told, I finished the overview in say 15 minutes and didn’t know what exactly to do after that. I was in no mood to implement any of the HTML5 features in interviewstreet now (Btw, if you don’t know, interviewstreet.com is my company where we simplify the hiring process by assessing the candidates using automated tests.)

Then for the next 15 minutes, I started playing around with some of the demos at http://html5demos.com/. They are cool, but nothing new or mind blowing. We’ve already seen these stuff in flash and javascript using libraries like jquery, prototype etc. So, now the browser gives native support to these features. What’s the big deal?

I guess unless I create something using HTML5, I’m never gonna realize it’s true potential. I started with the canvas tag.

So, after loads of google search, copy pasted code and a few cups of coffee, here is my first HTML5 game. This is the screenshot of the game.

You can play the game at http://sp2hari.com/html5/burst_the_balloon.html.

TODO:
1. Some sort of levels in the game, where you are taken to the next level once you shoot all the balloons.
2. Display score, time taken and other details.
3. Two player game, where two players can shoot arrows from both the directions and one who gets the highest score wins.
4. A special balloon which gives you more score if you shoot it.
5. Better animation for arrow and balloon. Right now, the balloon looks like an egg.

Anyone interested in playing/learning with HTML5 can take up this code and implement something from the above TODO list. Do mail be back at sp2hari AT gmail DOT com if you implement anything or even if you have anything to say about this game

More than an year since I’ve quit IBM and into startup world. Too many things to write.. Things moving at such a rapid rate that if I write something today, tomorrow it might not be valid or correct. Makes me stop and think at times whether what I’m doing makes sense, is it the right path or am I yet another crazy person.

Advices are something whichi I’ve got more than I’ve wanted in this one year. Way too many advices. Contradictory advices. Person X says do this and Y will say don’t do this. Among all the advices I’ve got till now, I remember just 3.. These 3 are now a part and parcel of every small decision I make and are a part of my system.

Swaroop is currently working in Infibeam and has worked in Adobe, Yahoo and his own start-up. He blogs at http://www.swaroopch.com/. We met swaroop when he was just about to join infibeam and the main goal was share his experiences and basically to get some start-up gyaan.

He started the whole conversation with the following disclaimer

Hey, I'm going to say a lot of things which I've learnt from my experiences. I might give loads of experiences. But don't take them for granted. Never apply them directly in your business. Anything I say, please use your brain and make sure you check whether it makes sense for you to follow it or not. We met a bunch of VC's and every VC suggested a change which changed the business by like 10%. After meeting 10 VC's and implementing a whole lot of changes in our product, we came to realised that we've changed the product by 100% and what we have now is not what we started to do.


If you don’t consider that as an advice, read it again twice.. thrice.. maybe 100 times. I did the same mistake. I didn’t realise I was doing it before I met him. I knew I was making tangential changes. But what I forgot was these tangential changes add up and change your product whole lot.

So, every customer or client has his/her own requirement. Before we met him, we used to put a night-outer, finish the feature for that customer and try to make him happy. What we failed to think was what will happen beyond that customer.

Now, things wait. We take 2 days even to start a new feature. We have a build cycle, where in our new features gets pushed to production only on Mondays. So, even if I code something fast and quick, I have to wait till Monday to get it to production. So, I take this time to think more about the feature and this helps us a lot.

The second advice is again way too obvious and I can’t believe I started following this only after I heard it from someone.
Nithya Dayal is the CEO of Muziboo.com, where you can upload music and share it with friends. I happened to meet her in Yahoo Big-Thinkers event and we were talking about Product Management and what role does she perform as a CEO of Muziboo. So, she mentioned like ”

As a developer, it is easy to get excited with new technology, tools, products etc. So, what a developer does and what you've planned for the product might diverge at times. It is my job to make sure that what has been planned is executed. If you don't follow your plan, then the hours you spent on making the plan goes down the drain. So, either you shouldn't plan or if you do it, you should execute it

How this helped me?
Lots lots and LOTS. I stopped doing multiple things at a time. Both in my desktop and my mind, I had only one thing to do. Stopped switch tasks and realized that “Developers are monkey minded.” Even now, it is so damn itching to work on cool stuff like scaling, mysql replication, multiple servers and other stuff. But the plan in the board says “Get 10 paying customers and then think about anything else”. So, back to the small small features customers asked for and trying to close the deal than work on new cool features.

Last one, is from Mukund from Buzzgain. If you’ve by any chance met him, you will become his fan. Simply awesome guy whose enthusiasm level is never less than 100% . So, we showed him a demo of our product and he liked it, suggested changes, sales tips and while we are about to leave and say bye, he asked us a question.

"Why didn't you ask for more contacts from me? You SHOULD ask that. Do you remember the binary tree you read in college, that is how you should do sales. Every person you meet ask him 3 more contacts. ".

I was speechless. I was SO dumb that getting more contacts from him never even occurred to me. Again, a simple advice. I’ve made it a habit now.

I hope following just 3 advices over a period of 12 months is not so bad after all.