Archit Baweja's Blog

The Amazing AWS and Friends

2011-09-20T20:14:00.000-07:00

So after much time, I finally decided to take the plunge and try out amazon web services. There were a number of factors that finally pushed me over the 'edge' and give it a shot including

The upcoming release of my first Android App (more on that later) that I've been working on for the past 3 months.
My experiments with Drupal v6 (finally decided to wipe the dust off my Drupal 6 books)
And here's the kicker, Amazon decided to make their 'micro' tier for the EC2 webservice free for uptil an year.

A little background on Amazon Web Services (AWS for short) is in order. Amazon Web Services are a host of services offered by Amazon Inc. which allow the consumer to use their infrastructure to do various hosting-like (note, 'like', its not exactly a traditional host provider) services. The AWS actually consist of very many services including a service to run virtual servers for temporary or permanent periods (specifically known as EC2), storage services (Amazon S3) etc. From my limited experience there are more on the way. To be clear my experiments have been mainly with the EC2 service right now.

At its basic the EC2 service provides the user with a virtual server in the cloud (queue divine music) which can be used for any kind of computation. The virtual server is a complete OS environment, and the EC2 service supports a host of OSes, mainly windows and linux flavours. You can use this virtual server to accomplish many things including

Host a website
Use it to run a temporary computation for results (simulation etc)
Any other need you might have for a temporary or permanent computer within a few clicks.

Again my main goal right now was to get started with EC2 to host a website.

To any new user I would personally recommend reading the numerous other blog posts and youtube videos which go in various amounts of details in setting up an EC2 instance/virtual-server. If you feel up to it, you can always read the official documentation at the AWS website, but I feel the tone of much of the documentation is quite PR oriented and can get in the way of trying to get up and running. If you're like me you like to dive right in and experiment I'd stay away from that initially. Some of the more useful links I found on getting started were

An awesome article (details) on quickly getting setup with a LAMP stack on Ubuntu Linux with Drupal 6, http://undefinedvalue.com/2010/11/12/setting-drupal-ubuntu-1010-ec2
Couple youtube videos that I bumped into including http://www.youtube.com/watch?v=ZAB8wCg9MyE&NR=1

Again the point I'd emphasis is to have the official documentation open for later, but instead go with a standard google search to get started. I won't go into the details of a setup, because there is enough web resources on it. They will also have help you get over the flood of new terminology that might hit you like a truck if you go straight for the official documentation.

Overall I'd say my experience was relatively straightforward and smooth. Despite me treading slowly, I was able to get an instance up and running pretty quickly. My previous experience with the LAMP stack in general of course helped me greatly.

Possibly the most novel thing about AWS as a whole is their billing model. You are not billed like a traditional host provider. There is actually a very minimal fee for the EC2 instances, however the majority of the billing is done based on 'use' measured in terms of bandwidth, data transfer etc. I feel that this provides a much better billing model for the end consumer; you pay as you go. Even evaluating just EC2 as a website host, it makes sense for a lot of personal or low traffic websites to consider EC2 as a platform, so that you don't waste money and bandwidth which you might in traditional website host providers. And for periods where you do get more traffic, you end up paying only for those periods.

In the end I'd recommend every developer go ahead and try it out. The Cloud, after all, is the future.

Talking Movies: The Day the Movies Died

2011-03-14T08:00:00.000-07:00

In the recent issue of GQ, Mark Harris wrote a brilliant article on the decline of movie culture in America. Ever since my first "english-pichure" in Chanakya Puri Cinema (R.I.P) back home in New Delhi, I have always been fascinated by the silver screen art form. The stories, the characters that comprise them, the effects and all the ingredients that make for a movie come alive (and I don't mean 3D!) are an integral part of a new frontier in story telling in our culture (well not as new as video game narratives). However this art form is under danger from its own success and its big blockbuster nature, the suits ready to trivialize stories and their characters for the next big buck.

The main theme of the article dealt with the brand-ification of movies. The central idea of his argument being that with more and more money on the line for each movie release, movies are being viewed more and more as just another product which is made for public consumption. And the main rule in such ventures is profit. Movie producers are only willing to bet the massive amount of money if they can have better guarantees that their returns will be safe and within expected figures. Brand-ification helps in this case. Brands are safe. Brands allow consumers to relate to past positive experiences and make them keep coming back for more. But where previously the brand was the artist (the script writer, the director) it is now the art it self (prequels, sequels, blah blah blah). I can definitely relate to that. Previously when a new movie came out and say for example had a certain director's name attached to it, movie goers would be urged to go for them. Nowadays that link to new movies is made based on the movie itself. Pirates of the Caribbean 1 was good, great so 2 and 3 have to be good too! I use the example of the Pirates franchise because by the end of the 3rd installment, I felt the story had been total shit. The movie was just trying to sell based on the A-list actors and the inter-character chemistry. There was not much to the story itself. There was nothing new to the characters.

This only smells trouble for the artist movie maker who's vision is sidelined for mere profit. No longer are movie producers willing to take risks in the name of the artistic vision. For example, more and more movies are being made based on published literature. To the average consumer it may seem like the movie producers are bringing their favorite book to the big screen, but the cynic in me cannot help but think that the movie producer is just going after a tried and trusted story; a book that has been published and its success ratified. Only then will the big movie studio's be ready to commit.

The argument put forward by Mark Harris, also lead to a possible reason for an interesting phenomena I have noticed in the content of movies from the 90's to the 2000's. There are far lesser cuss-words, F-bombs in regular movies nowadays then back. Take for example the Die Hard series v/s Bourne Identity. The writing for such an action film should have room for a few choice words here and there, but there is a marked difference. I think the reason behind such seemingly minor details is that it allows studio's to get favorable MPAA rating. Favorable in terms of the size of the audience that is allowed to see. This directly leads to more sales and more money for the studios. All the while the content of the movie itself suffers.

While I feel the article feels resigned to the gloomy fate for the movies, to go so far as to even lay some blame with the audiences (with which I do agree to some degree), I personally am more hopeful of the future of movies. Audiences do crave new and fresh stories and characters in movies. Where the big studios failed in their commitment to original story telling, I do see the rise of the Indie movies. Recent commercial success and industry recognition of indie movies from the likes of Juno or 127 Hours (debate bait ;-) ) my faith in the audience's taste for movies remains steadfast. I'm personally very grateful being a Philadelphia resident that in this city we have the wonderful chain of Ritz movie theaters, traditionally known for showcasing indie movies. Even over in Bollywood I see the success of Indie movies like Peepli Live and Udaan, it only goes to show that the movies are not destined to a fate of fastfoodnation-ization (tm).

[1] http://www.gq.com/entertainment/movies-and-tv/201102/the-day-the-movies-died-mark-harris

Progress meters in web design. Do we really need them?

2009-07-31T07:35:00.000-07:00

Today I'm going to talk about some of my experiences with using the excellent Gaia Flash Framework. Along they way I figured I'd describe some of my experiences with not only the framework, but also the Flash CS3 and AS3 platform. I'm a relative new comer to the Flash Cs3 platform and even though I've had (imo) enough programming experience with other platforms, there were a few things in AS3 that caught me by surprise (good and bad).

Over the course of past 2 months, I've had the opportunity to work on my company's own website. Its been quite a blast. Part of the fun for me on this project has been working with a completely new platform that I'm not familiar with. Learning a new platform, although has its lows (more on that later), for me there has always has been that attached excitement in learning something new and looking at things from a different angle. Plus in my experience, learning to program in a new platform is like maths, its better if you practice it along. You pick up more and learn as you go.

So after setting up the initial website, I came to the point where I had to set up a whole section to display the various portfolio pieces Odd Graphic has done (print, media, multimedia, websites). Now initially I had a coverflow-esque idea of how things would look on the portfolio section. What it boiled down to was a series of images (png, jpeg, pdfs) of various client work Odd had done in the past. Instead of picking up a pre-made Flash component, I decided to program from the ground up, in the hope that I'd learn a few more things.

Now while working on the portfolio section, it dawned on me an interesting thing about loading progress bars. In this case basically when the user wants to view Odd's previous work for a client, the website has to load 'n' number of images and display them as a slideshow (user controlled or automatic, thats a different matter). Most websites (or implementations) of such a feature involve the user being displayed a progress bar (or a meter to be precise) while the flash player loads the images to be shown in the slide show. Once all the images have been loaded, it displays the Flash component with all the images (items). And along the way as I was checking out examples of using customized progress meters with Gaia, I noticed that designers put in effort in designing the progress meters themselves. I got a sense that progress meters had become an accepted part of rich websites, an almost integral part.

Now being a programmer I know that is a good starting point, but it isn't the end. I know there are issues like network latency and usability concerns, which is why progress meters are in use. However having the user wait while all the images load is also a usability concern. In our world of concurrency, I think we can do better by following a alternate programming work flow. To summarize, the way programming pattern followed by most websites is...

Show progress meter
Load all the images in the background.
Once all images loaded, start the slide show.

What I suggest would be a bit more mature. Something which uses concurrency to download the images all the while displaying the images that are already downloaded. The benefit of such an approach is that the user gets to view the images as soon as possible. Most slides hows any way have a certain time period (a few seconds) before they move on to the next item. So why not grab the user's attention asap and get him hooked while the rest of the items are being loaded on the network. The alternate pattern would be then...

Show a progress meter (now hold on ;))
Load the first image.
Hide the progress meter, show the first image, and start a concurrent request to load the next image in the background.

Now a progress meter is needed still, in case of emergency (big spikes in network slowness etc), but in most cases, the user would be presented with the first slide show item while the website (with just a little extra work) would end up doing loading the rest of the slide show items, without the need for a progress meter.

What do you think? Do you think we as website developers are just sticking to tried and tested solutions when we can do better; both from a programmer point of view and user experience point of view? I'd like to hear your opinion.

Fortunately, based on what I discuss here it is very convenient to follow such a slide show loading strategy in flash using the Gaia Framework. Stay tuned till next time when I show some real code with respect to this problem.

Back and about

2009-07-17T06:54:00.001-07:00

Phew! So after some random clicking and jumping across web pages I have finally managed to recover my old Blog account. Its been a while since I kept a blog for any purpose, but recent changes in my life have lead to me to start one again (I used to have an old one about random programming experiences while I worked on various open source projects @ advogato ; on closer look, wow thats even older than the first blog post here).

So whats the purpose of this blog? Well I recently got a job @ Odd Graphic as a multimedia director (read: web apps and website developer) and I'm having a blast. Besides the obvious monetary reasons, its finally good to be out of college and in the professional scene. Its good to finally put into practice all those 7 years of college. Time to see if it finally pays off. Being the main and only programmer at Odd Graphic, my job mainly involves developing and maintaining websites for our various clients. This can include using a host of development platforms and hosting services including Rails, CakePHP, Adobe Flash CS3 suite, Actionscript 3 (I kind of decided to forgo AS2 completely, more on that some other time), CSS, XHTML, Javascript etc (you get the idea). For older clients the decision to use a particular platform have been made already. For newer ones, it depends on the requirements of the clients. Uptil now the niche in which Odd Graphic operates , the clients don't usually have very many technical requirements for the work they require done. All they care about is the job getting done and on time. Of course, being programmers we know that requirements for a given project aren't always complete; especially at the scale of projects I'm involved in. And then there are requirements like security, performance which aren't always stated upfront, but being diligent software developers I feel its our duty to keep them in mind these issues and be prepared for them if the client suddenly feels the need to bring it up.

From my perspective, I try to be pragmatic in terms of how I get work done. Firstly I feel I'm past the point where I should waste time on religous wars about tools, services and platforms, but instead focus on getting the job done. As an example, in my nearly 2 months of work at Odd Graphic, alot of my work has involved making websites for clients. Some are for internal use, but mainly a lot of the website work is for attracting further customers for their businesses. Websites like these should be designed to be viewable by the largest audience available. Sure the whole line of Internet Explorer's has the dodgiest support for XHTML/CSS standards. Sure it sucks when you develop and test a website on Safari, Firefox, and IE7,and when you go in for the demo, only to be faced by a blank screen (only to realize that the largest audience of the website is going to be using IE6 ). But despite the pain, I feel its my job to get the job done. Occasions like these also give me a chance to reflect on what changes I could have made to my toolset to minimize browser irregularities. I know that this very problem can be remedied (if not by Microsoft :P) by a very fundamental concept in most programming languages, abstraction. If only I had used some great freely available libraries, these browser irregularities could have been avoided.

As you may have already guessed I'm going to post about my day to day experiences and reflections at this job (amongst other things). From daily technical issues encountered in programming languages to various tools I use. I'll also talk about any blogs I avidly follow for work or personal reasons and about articles I read that I find interesting. For example I've lately picked up on the Adobe Flash CS3 and Actionscript 3 platform for various projects that demand it. As part of using these technologies to develop websites I also came across the excellent Gaia Flash Framework, that makes developing websites in Flash quite painless (well more often than not).

Since this is my first and only blog, I don't feel the need to start more blogs for various other things that go on in my life. I'd rather have just one place to dump my thoughts and ideas about everything. From cocktail recipes, to random attempts at poetry (haiku's anyone) , they will all be here.

And so with a giddy feeling in my stomach, and trembling fingers as I drag the mouse over to the publish post button, I bit you all adieu. Stay tuned for more in the coming weeks :)