Continued funding an IR is largely a matter of suceess. The library is willing to devote resources to initiate the project and can probably maintain the project for its first year with its current hardware. If the project were to become truly sucessfull every facet of the project would have to be evaluated.

The two main determinants of cost are scope and technical execution. Project cost remind one of the old saw "Fast, Cheap, Good; pick two".

The technical requirements of the IR are continuing to change but most of those changes are leveraged by the use of open source software.

In many situations complexity can not be reduced but it can be distributed. If the user takes on more of the burden of using the system then the system can be simpler. Unfortunatly most online systems have to be very user friendly or users simple won't use the system. Hopefully we will be able to find some balance on this matter.

Operational costs are often a function of use; the more users the more support staff necessary. If we're lucky enough to create a usable interface and can automate much of the submission process operting costs of the repository should scale.

Were we see a more linear growth is in supporting faculty, staff, and students in creating and formating the objects they wish to submit to the repository. These costs will definatly be beyond the ability of the library to support.