With today's technology it is very easy to collaborate with people around the world (I totally recommend Vyew for that). When you collaborate with people from different countries, are there any rules for who pays the bills? The expenses could be conference fees, attendance fees, journal fees, etc. For collaborations we use Hardy-Littlewood, but is there any such kind of rules for paying the bills? Or is it just a previous agreement or consensus between your coauthors? How do you normally approach the subject to your fellow coauthors?
In theoretical computer science (which I work in), the most direct answer is "What bills?"
For remote/electronic collaboration, there really are no significant costs. Each collaborator uses their own computer equipment and their own internet access. They install LaTeX and Skype and svn themselves. They typeset and illustrate their own papers. They each buy their own espresso at their own favorite coffee house. There are no fees associated with submitting or publishing papers.
On the other hand, if you actually want to travel, either to give a conference talk or to work in the same physical location as someone else, there are real costs. For conferences, usually each person (or their grant, or their advisor, or their department) pays for their own travel, lodging, and registration fees. For collaborative travel, every trip is different, but some common protocols include:
- The visitor (or their advisor/grant/department) pays all costs for their visit.
- The visitor (or their advisor/grant/department) pays for travel, and the host (or their advisor/grant/department) pays for lodging.
- The host (or their advisor/grant/department) pays all costs for the visit.
I have used all four of these arrangements (and various ad hoc mixtures), as visitor, as visitor's advisor, and as host. For some visits, I've had small joint grants specifically for collaborative travel; in others, I've used travel money from a larger grant; in others, I've paid (either for myself or for my guest) out of my own pocket.
In short, F'X is exactly right: It depends. Obviously.