I think there should be a fixed maximum time limit of completing the process of either accepting or asking revision or rejecting. In either case the journals should put the abstract of the submitted manuscript on line open access to avoid losing priority of independent works. Recently i reviewed a paper after two revision I have agreed to accept but the paper is yet to come on line. In between I have received another paper  for review from another journal with lots of common work ( with the one that I have already accepted) by a separate set of authors!! Now had the abstract of the first paper here  had been made available as soon as it is submitted this could have been avoided. More over the authors of the later paper could have contacted the formers to avoid such situations.

Subrata -

You certainly have touched on an important issue (or two) that needs to be addressed, though the ' solution' may not be as simple as one might like. As a referee/editor I often have felt rushed because I did not want to hold up an author or authors publication, but then, I did not want to provide a faulty review. (I just received a journal paper to review, with no warning - though that is a streamlined journal in general; and at the same time, a dissertation I was told I would receive, but had to keep asking where it was, as I was worried about the schedule.) On the other hand, I have submitted a paper to a journal notorious for taking a couple of years in their process, yet in spite of the time they take, I received a comment from a referee that was patently incorrect. How could they be so slow, and still be so sloppy, and yet have an excellent reputation? (They are good, and I have refereed several times for them.) The point is that it may be more difficult for a journal to do a good job and be timely, than would be intuitively expected.

So what are the solutions? First I think that some delay may be involved in passing manuscripts between editors and referees, and other communication delays that a journal should try to streamline. Another solution is one of which I took advantage: publish much of your work quickly in conference papers and/or a more lightly edited journal, and availability can be enhanced by posting papers on ResearchGate.

You noted a third possibility of posting abstracts early. That is an interesting proposition, though I think it needs to be considered carefully. A colleague had a journal paper being held up, where he had an idea that has since become a largely used survey statistics tool. While his paper was being held up, I was told that another author, apparently with ties to someone in this process, had his paper moved forward on the same topic, in another journal, and published first. If abstracts are posted early, might not this sort of thing happen more often? Unfortunately there are such shenanigans. Your solution may be derailed by problems of professional ethics.

I don't know the answer, but delays in publication do seem a problem. As indicated above, I suspect that wasted time when shuffling manuscripts between people and/or communication issues may be part of the problem that might be most successfully addressed.

Also, your idea regarding abstracts deserves consideration. Perhaps some modifications to this idea may be discussed. (One problem may be 'pilfered' ideas, but on the other hand, another may be someone, by submitting an abstract first, trying to stake too wide of a claim on future work/results.)

A journal should help people publicize their work so that not only can the author(s) add to their credits, but others can build on that work. Cooperation is often far better than competition, in my opinion. So it may be helpful for everyone if journal articles could be published under shorter schedules.

Cheers - Jim 

Thanks Jim so such elaborate discussion with points to be seriously debated. Yes I can see the loop wholes in uploading abstract immediately. the current model need to be revised to see that the time take from submission to decision is reasonably small.

I want to share a trend which is also putting question marks on the seriousness of a journals in selecting reviewers. I have found in more than one occasion that papers are sent for review to some one who is actually a Ph. D. student and have just published a paper as a coauthor. This is in general may not serve the purpose. This would not have happened if the published paper carry short description of the authors. As done by the IEEE journals.

Subrata,

I agree with Jim that it is very much helpful for everyone if journal articles could be published under short time schedules. I think there should be a maximum time limit of no more than 3-4 weeks for completing the process of peer review of any article submitted for publication in any journal until online publication. Before the year 2000, all communications were done by post and the preparation of the manuscripts until getting them published on papers used to take a very long time. However, although communications and processing became so fast, many journals still delay the publication.

Quick would be nice, but 3 to 4 weeks, in my opinion, may be too ambitious. Quick could help eliminate frivolous demands by referees and editors, so i am in favor of that, but on the other hand, conscientious referees and editors can be overwhelmed (as well as there being too few referees and editors available) such that 3 to 4 weeks does not leave - in my opinion - enough time, especially as in a case I am editing now for a quick online journal, where I am having to ask the author some questions so that we don't post something that is just wrong. I just wrote to the author that I needed a certain explanation before the paper could be posted for others to review.

Work on which someone has spent months, based on perhaps years of experience, cannot always be quickly examined (especially if the referees and editors are busy with other activities). Such work needs to be separated from a paper quickly done in error by someone who is not really an expert. Sometimes that takes a bit of an examination, in statistics anyway, as an equation, much less pages of them, can easily conceal an important flaw. I am an easy referee, but I had to reject a paper not long ago, for a journal with a substantial impact factor (for a statistics journal), because it turned out to rest on a false assumption. That is not always quick to 'dig out.'

http://ijaridea.com/Peer-Review-Process gives you appropriate answer