Reflections: Tech

Open Source: An Analysis

2008-06-22T12:36:00.000-07:00

I find open source to be an interesting phenomenon... I love the concept when it is at the realm of learning and I undestand its power in growing. It can help something grow at a massive pace.

I watched a video recently (as part of a course work); the presenter was saying that Lego had 3 engineers working for 7 years - 21 person-years before it got Lego Mindstorm out. Within a few weeks of the release thousands of hackers around the world made their own creations, and demonstrated that they were better than Lego's creation. (I abstract it to creation because I am not sure what exactly it was that they did better :) )

Open source could work very well for high speed of technological progress; some one works, puts their work online; someone else starts working, looks at the first persons work, learns and understands from it, pushes it higher; more people keep doing this, the free exchange of ideas can lead to great ideas and fast progress.

BUT, here is what puzzles me, where is the money? (not to sound bad, but isn't that the bottom line?). I was talking to a friend, he asked me how is this different from people posting videos on youtube - they don't get paid.... I understand that both are similar in that they may be forms of self-expression. But what the programmer does is worth something, the youtube user's video may not be worth as much... (this can be a good point of debage, may be a topic for another post ;) )

May be it is just the passion/challenge of the hunt that the programmer feels is worthy enough. Or in a more human form, the recognition that may be obained from an online community may please them and make them feel is payment enough. Its interesting... To get answers to these questions might mean we need to explore deeper into how we actually think... and I feel a little tired to do that now. :)

Are we blinding ourselves by this short term (or is it eternal?) recognition into changing the dynamics of the software development field.

If this goes on, I see that very good software becomes freely available; company's may not have the need to buy software. It is just the service (support) that they may need to buy. Some open source software have already adopted this model. To provide support, do you need to be a programmer? Mostly no, but then there will always be a need of central control in such a model, and there will have to be programmers in its center (in the case a specific customer has encountered some bug and needs an urgent fix, you can't rely on the community to deliver with the speed needed by the customer). To do the function described above, these programmers would have to be highly skilled (or specialized). This eliminates middle-tiered programmers in this model.

Through the laws of supply and demand, let us say, we reach a point that there are many many programmers (due to the mass spread of knowledge) and the eliminated need for middle-tiered programmers - the supply greatly exceeds the demand. Given that programming has just turned into an unfavourable career path, there would be a slouch in the number of programmers. This autocorrection would keep only those programmers who can deliver in the field (the high calibre programmers/specialists).

As time passes, if more people move away from programming, given the bitterness it would have presented in the past, the supply would start to fall relative to the demand. This would become the 2nd birth of more programmers. It may be reasonable to say, that the lessons from the past would keep a good equilibrium from this point on (of course this goes into predicting the very long term, which would depend on many other factors at that time).

There is a lot more that can be said in this topic ... and I feel tired to continue on. Let me stop here temporarily and say "to be continued"

Tip of the day?

2008-06-21T21:35:00.000-07:00

I was hearing a Tech podcast - WebDevRadio - the second episode (the episode was from 2 years ago); the episode is about Web Optimization. Here are some of the presented tips:

Send back less HTML to the client! Strip whitespaces, extra carriage returns, and HTML comments - they do not help the user viewing the page. If you have a dynamic site, you can use functions to do these automatically. So you do not need to change page by page.
In PHP you can use output buffer, store the output, in the end run a function that would do all the above clean up and then gives it out.
This reduces bandwidth usage. This could help high traffic sites quite a bit.
Have external links to JS files rather than inlining it in the HTML page; this reduced bandwidth usage. The linked file gets cached by the client browser, and gets used on successive requests.
Compress your Javascript. There are code projects and functions that do this.
Using Gzip compression for output data - the client sends the data that "I can accept Gzip compression", the server can now send the gzipped data. The bandwidth usage reduces BUT at the cost of CPU time ... this becomes somewhat of a judgement call.
Cookies: Use as few as possible. Possibly just 1. If you had 5 cookies, every request to the server would be pushing the 5 cookies back to the server. Say each cookie was 2K, then every request would be sending an additional of 10K of data.
CSS: Using CSS effectively (e.g. classes) rather than HTML for formatting (think about repeated use of " ... ") could reduce the page size of HTML pages.

Software Development Life Cycle

2008-06-03T13:57:00.001-07:00

Here is a funny picture of the stages of the SDLC. I have seen this before in an email and also posted in this blog at agiledeveloper.com

Funny stuff =)

disclaimer: I did not design the image, so I don't want to take credit for it. I found the image in the above site (I did not want to hotlink it and make them pay - as in bandwidth - for the content viewed in my site)

New Blog

2008-05-19T19:20:00.000-07:00

Ai-Projects.info has been refactored. The site is split into 4 zones now:

Personal Blog
Tech Blog
Resume
Contact Page

This is the tech blog. I will post any technology related things that I want to talk about here. I include programming when I say technology, so this may include algorithms etc...

The old blog (which leaned towards being a tech blog, though not specified as such) can be found here - http://www.ai-projects.info/SphpBlog/