Anyway, on top of my own issues, I was reminded of the perils of length/reference restrictions by an email from Jean-Michel Claverie from CNRS. In the email he told me of a situation involving a recent paper in a high profile journal with a name that begins with the letter N. This paper did not cite some highly relevant earlier work of Claverie's in PLoS One and when he wrote to the author to politely point this out, he was told that the reference was basically removed for space reasons. I have seen this happen many times with a variety of journals and the explanation for some lack of reference to relevant work is always something like "oh yes, of course we knew about that, but had to leave it out for space reasons" or "well, you know, they only allow 30 citations, so we had to leave some things out".
Sure, in the past, when printing articles and keeping track of references was difficult, this may have made sense. But a HUGE part of science is giving and getting credit for work. And thus it baffles me why some journals enforce strict restrictions on the number of references allowed. Basically what this says is - it does not matter what type of work you did - it could only possibly have been built up the work of (insert # here) previous studies. This is just wrong in so many ways. One option to solve this would be for these journals to allow expanded reference lists in online material - and for these lists to somehow get picked up by citation indexing systems. But this is something they need to solve. And until then, people should be aware that by publishing in such journals you may indirectly be doing a disservice to the people whose work contributed to your own.