My problem about writing a technical/research report/paper
I suspect you are finding this difficult because you are trying to write in an academic style, and are judging yourself as you are writing. I think the secret to easier writing is to spend as little time as possible "writing". Instead, spend your time jotting quick emails and editing stuff. Let me explain what I mean.
Get together with someone who knows nothing about your research, and explain it to them. Record yourself, or at least take notes on the main points that you address.
Now write your explanation down, but informally, as if you were writing an email to a friend. Better yet, make it a real email -- that will help you keep it informal.
Also gather any other emails you've written to your supervisor, or colleagues and friends. (If you're not already doing so, get into the habit of writing those emails regularly, while ideas are fresh in your mind.)
Take those emails, and paste them into an empty document. This will be the starting point of your paper.
Now rearrange the information as needed, to make it coherent. Everywhere there's a gap, stick in a message like NEED TO EXPLAIN X HERE.
At this point, you're beginning to have an idea of the structure of the paper. This would be a good time to write an outline, to see if you've left anything out. This outline goes into the document along with everything else. (You can delete it before submitting the paper.)
Now fill in the gaps. But write informally, as if you were writing an email to a friend.
Now go through your document, and make the sentences more formal. Don't overdo it, though. You don't want your writing to be stilted; you want it to sound natural.
Congratulations. You have the first draft of your paper, and you barely did any real "writing"; it was mostly just writing emails and editing.
It's much easier to make informal, but clear, writing more formal, than it is to make formal writing more clear!
Just some thoughts that occur to me while reading your question:
"Write down and make it as long as you can and we will make it short later"
To write well is to make every word count: to not waste a reader's time. Length should never be the goal. Say what you have to say as concisely and clearly as you can. Do not get into the habit of writing "filler".
Even after doing every thing by myself, I feel that I can not describe it in writing. While if it is about describing it orally I am confident that I am much better than many.
Writing well means keeping the reader in mind: what will capture their attention at the start of the paper, what they will need to understand before the next section starts, what are their expectations for what is going to come next. As such, a good starting point for writing is to consider that you are discussing the paper orally with an imaginary reader you meet over coffee. How do you get them interested in the topic you want to discuss? How do you convince them of the novelty of your idea? How do summarise your unique contribution? If you had a pen and paper to sketch some ideas down, what would you sketch?
If these are steps, what are these?
First you need an idea of the scope of what you want to write about.
Second you need a frame for the paper, typically the section headings, a rough idea of the goal of each section, how they will fit together, what the reader will learn from each part.
Third you may need a detailed frame for the paper. For example, I have given students headings (as comments in the file) for each paragraph. In the introduction:
- "What is the setting/area?"
- "What is the specific problem?"
- "Why is it interesting?"
- "What has been done about it before (on a high-level)?
- "What is our unique perspective for solving the problem?"
- "How will that be achieved concretely?"
- "What are our contributions / How is the paper structured?"
Some paragraphs can later be joined together, as necessary. Having this framework for (at least) an introduction in mind, I find, helps inexperienced writers get over the "cold start" problem. (More experienced writers may not need to follow this structure.)
Then you start ... at the start (well skip the abstract ... do that last). Take it piece by piece. If you feel you don't have enough information yet to write a part, skip to a part you can write and come back later.
Write your abstract and conclusion last. Make your conclusion reflect what was promised in the introduction (but now providing concrete details, such as a summary of results).
Many times I have got my office reputation destroyed because of going blank even after many days/weeks given to me for writing. I want to get rid-off this problem and want it never happen again.
Sitting there frustrated isn't going to help. But even though I would consider myself fairly experienced at writing, I often have the same problem.
Each paper is different so if you're not sure about creating the outline, you have to talk with someone else to help you plan out the paper/report or even just to get your own thoughts straight.
Even if you're not sure what to write, just write something. Most papers will go through several drafts so the first draft doesn't need to be perfect ... each draft just needs to be an improvement. And a first draft can help you get feedback from others to point you in the right direction.
First, "Write down and make it as long as you can and we will make it short later" is probably one of the best advices you are going to get in your entire life in terms of technical writing. As you progress, writing large text is always the easy part. Writing a self-contained text with the required theorems, proofs, results to fit within a conference / journal paper page limits is the real hard part.
You say you do not know what to write. Then you must write everything. What is your input data? High level description and in detail stats. What is your output data? Again high level description and in detail stats. What do you want to do with your methods? What are your methods actually doing? And how are they doing it? Again top-to-bottom. A high level description first and then move into extensive details.
Also compare with previous methods. What do previous methods do? What are their strength and weaknesses? What are YOUR strength and weaknesses? For what datasets your method performs best etc...
Go into exhaustive detail. E.g., in a CS reports you may say "we use Java and Hibernate on a Intel i5 workstation running CentOS with 6Gb of DDR3 RAM."... Initially do not omit anything. Write anything relevant to fill the pages. As you become more proficient in writing, you will see that you may distinguish what is really important to say and what to skip. Do not overthink. Start writing, even it initially seems not good enough. There is no need to be perfect initially. Once you have some pages written, come back and edit. Repeat the cycle with every couple of pages added. Soon, you will find that filling the pages will not be a problem. Trimming unimportant things from these pages will be the real challenge. Good luck!!