This question follows from a MathOverflow question, "Why should one subscribe to print Journals".
It seems obvious to me that having print journals in a library is beneficial. Yes, Arxiv, MathSciNet, Blogs and lecture notes by Mathematicians, Math Overflow, Wikipedia and Scholarpedia all of these have been extremely helpful in dissemination of research mathematics. Amongst other things electronic copies of articles helped immensely in increasing accessibility.
Most journals these days have an online version. Which probably only a few would argue is not beneficial.
What I can't get myself agree is the opinion that one should stop subscribing to print copies of journals altogether. The arguments against subscribing both print and online I have come across are cost, space constraint, and redundancy.
The difference between online and print+online versions of journals are often marginal. If space constraint is an issue then one can argue against having a library as well. Why subscribe to the journals at all as most of the articles are available in the internet free.
What I can't fathom is the argument that print version of journals have become redundant.
I am asking this as in a discussion regarding journal subscription many faculty members expressed the opinion that we should stop subscribing to print version of the journals. How does one defend the case for need of print journals?
Am I overly emotional and just nostalgic for old times? So many times I chanced upon a result while browsing through the pages of a journal; sometimes relevant to my own area, and sometimes totally unrelated but so exciting that it got me interested in that area.
I am sure this issue or debate is not limited to my university and I am asking this question here hoping to benefit from the comments and thoughts you may share.
My question is what are the benefits of subscribing print version of journals, even if online versions are available. How it helps the research of faculty and graduate students.