Writing a research paper

This post is for my personal satisfaction, just like all my other posts. I have been working on my thesis on URL redirection and I have been "trying" to write a research paper to submit on Internet Measurement Conference 2008.
Writing a research paper is altogether a different field. I am pretty good at writing code, I write decent documentation and kind-of-ok tutorials on how to use stuff. But I really suck at writing a research paper. There are so many issues that needs to be taken care of.
The problem that I face is that, every time I make a revision to the paper, I find the outcome to be good and I feel that I am almost there. But then when I meet my professor and discuss the paper, I realize the paper is not good after all. There are so many silly mistakes and at a stretch, I am only good at two revisions. After that, I really need someone to pick up the mistakes. I guess the brain gets some what saturated and after certain point, we are no good at picking mistakes, unless you are experienced. After all this, I seriously think that I am not good for a Programming Writer position at Microsoft, which I seriously aimed for.
Moreover, I am applying for CPT work authorization, after which I might possibly and hopefully work for ImageRight writing ASP.NET software. By the way, I was informed by Cheryl from Javaground that I did clear the interview and they do have an offer for me. I was informed that Alex Kral, the CEO would talk to me about the offer. I am really excited to say the least but given the Californian conditions, I am again giving it a thought. But to be frank, I think that would be a totally cool job to take up and it would be a great learning experience. I would have to hear their offer and then make a decision based on that. But right now, I am looking towards the internship at Imageright. They have an internal project which I plan to complete by end of July. And if I get the CPT work authorization, I need to work on my ADO.NET skills seriously since it has been quiet a long time since I wrote any code ado.net. Frankly, I am not so good at ADO.NET. Let us see how it goes.
From all this, I really appreciate my major professor - Kang Li (http://cs.uga.edu/~kangli) for that he has been very supportive(though tough sometimes) in all aspects from the past one year. My first course at UGA was taught by him and since then I have been working with him. People at UGa, if you are reading this and if you are from computer science - do not be afraid of Kang Li or Lowenthal. They are all great professors but a little tough at times. I think the Dept of CS at UGA itself is the most friendly department I have ever seen.