There is a correlation (most good CS researchers publish in journals), but it is not as strong as in physics. Also, the lead times are much longer. In CS confer

There is a correlation (most good CS researchers publish in journals), but it is not as strong as in physics. Also, the lead times are much longer. In CS conferences typically play the role that journals play in most other fields. The fact that CS mostly uses refereed conferences as the publishing medium is an ongoing issue when comparing with other fields. One favorite anecdote I heard was the CS guy who was told off at a faculty review by a physicist for getting a "Best paper" award at a conference, he shouldn't spend so much time on frivolous presentations :) Unpublished claims of superconductivity and cold-fusion should be regarded as suspiciously as unpublished claims of proving P!=NP.