Adviser wants to be joint first author, even though I did most of the research

You write:

The only "effort" my adviser does is writing the actual paper

Well, in many fields, that would make them sole author. It certainly gives them a very strong claim to be first author, in fields where authorship is determined by contribution to the paper.

If you want to be first author, then write the paper yourself.

And if you can't, then don't expect first authorship, and be grateful for co-authorship (rather than an acknowledgement).

You've now added that you have done a bit of the writing, but your adviser has done most of it. So yes, they would, in many fields, still get first authorship, and you would get a co-authorship on that basis, assuming what you've written is a non-trivial proportion of the whole.

Azer89, Sorry this is stressful for you, most of us have had some sort of grad drama. I suggest you relax a bit about it. If your adviser demands "co-first authorship" I'd let him have it, for a number of reasons:

1) He is your primary job reference and entree into the academic world, keeping him happy is important. Your career will be better served with a good reference than any one paper.

2) It sounds like he has in fact met the standard for co-authorship, I wish my MS adviser would have anything to do with the manuscript!

3) Finally, and most important, is what kind of collaborator you want to be. Academia (and industry) is full of territorial, jealous types who are concerned only with their own academic "reputation". Don't be one of them. In the long run nobody cares if you are sole or joint first author on some paper. They care if you are a good collaborator they can work with. Take the moral high ground and get the paper out. If he's a real jerk you'll never work with him again, but you'll be the guy people can get along with.

Good luck!

Remote doctoring issues between students and advisors is always difficult, but I'll have a stab at it anyway.

In my research, My adviser only gives minimal advice that limited to the direction of the research and defining the problems. I can say I do most of the thinking, the actual solutions of the problems. Even at one time I needed to explain to my adviser how my algorithms work. The only "effort" my adviser does is writing the actual paper, since I'm still a master student with lack experience of writing. Moreover, I'm not a native speaker so I'm still trying to improve my English writing.

While your advisor may not be the most contributing person to this research, calling this "no significant contribution" may be too much. It sounds like he is doing standard advising plus helping you write the paper, which is certainly in line with what you expect from an advisor.

The actual problem is my adviser demands him as joint first author.

Yes, that is a problem. It's good that you push back here.

He told me that He wants tenure position and He needs my help so he can be an associate professor.

I don't quite understand that. For all universities I know, advising students is what is expected from assistant professors. So from a tenure point of view, advising a student to write an excellent paper is not much different than writing one yourself. Maybe he is overestimating the additional value he would get from being first author.

(Oh man, did he beg to me or what?)

Try to stay focused on the facts, and don't let it get personal. Nothing to gain from this.

He can be the joint first author but my name should be written first.

I am still confused about the notion of joint first authors (I always assumed there can really be only one, but apparently I was wrong). Anyway, I would say it sounds like an ok compromise.

I need to graduate In my uni, professors have total control to let students graduate.

That is indeed a practical problem in many universities.

A few concluding statements:

• Authorship really isn't something that should be negotiated after the project is almost done. As many members here never get tired of saying, details of authorship are best discussed at the beginning of the project, so that everybody knows where everybody else stands.
• From your short description, it seems to me like you are somewhat underselling the contributions of your advisor, even delivering some underhand blows in a few places. You are likely a rational individual - try to not be emotional about it, and then re-check what your advisor actually does for your project. Further, keep in mind that students tend to over-value the technical "doing" of the project and undervalue the "big picture" (deciding on a research project, defining research issues so that the outcome is both achievable and novel, etc.).
• Somewhat related: your advisor is not supposed to be an all-knowning, all-understanding supreme being. It is completely ok that you occasionally need to explain the details of an algorithm to him. That does not make him incompetent. You will for sure have to do the same thing with your new advisor.