FSP started an interesting discussion at her Scientopia blog on "pity PhDs". She is responding to a question from a PhD student about to start their 3rd year on a nonworking project. Now, given that this student still has 3-4 years left (or would in my field anyway), recognizing a potential failed project means they have plenty of time to regroup. My own project was only marginally successful at best, but that doesn't have to be a fatal flaw in getting a job, even in academia.
As I said in my comment, I would hire a postdoc with few (or one) publication who came highly recommended. I would be more likely to do so if I knew the recommender (and could guess at the likelihood of being snowed). I would not hire a postdoc with NO publications.
Students who plan a career in research KNOW that publications are key. As a PI, I want to know that someone can finish what they started, can write at least a little, and has gone through the process of converting lab work to manuscript. I do understand that not all projects are successful (which is why they call it research), but that is no excuse for having no publications in 5-6 years of grad school, especially if you plan on an academic postdoc.
My own PhD project was only marginally successful, leading to one publication that I submitted after I started my postdoc. However, like FSP's reader, I saw the writing on the wall. In my third year of grad school, I took on a side project that eventually led to 3 publications. Sometimes, things don't go the way you hope. I think this is actually a GOOD thing for a student, because it helps you learn troubleshooting and triaging skills. Unfortunately, even the very best advisor might not notice that a project has a fatal flaw until it is too late for an individual student. Anyone can fall in love with an idea or some lovely preliminary data and be unable or unwilling to respond quickly to a flawed research direction.
The thing is that no one cares about your career as much as you do. You need to be proactive, even as a student. If you think your project will not produce in time for you take have publications, YOU need to find a side project or two that will. For your side project, you can't pick something else risky and flashy--this is something that has to produce something quickly for you. You also need to go over everything you have done for your main project and see if any of it can be put together for publication. If you are funded by a particular grant, you (and your PI) are on the hook for that project. But most advisors won't care if you do something else on the side, especially if it doesn't require many resources. If one of my students came to me with some really interesting data, I would encourage them to keep working on it and help them get what they need to be successful (as long as whatever they were supposed to be doing continued to get done).
If all else fails, and you do end up with a pity PhD, I can say that the people I knew at PhD U who got pity PhDs are all working in industry quite successfully right now.