R help - Jan 2016 - parameter constraints in glm() and Bayesian version

Hello,

I'm a mathematics student at Ulm University and currently I am working  
on my bachelor thesis about a Poisson regression model.

For this, I am using the function glm () in R which is working very well.
But still I have two questions to improve my model and I hope that you  
could help me:

(i) Is there a possibility to set constraints on the regression  
parameters in glm() or is there another function in R?
Specifically, my paramters should be constrained to be positive as  
negative parameters wouldn't make sense. How can I do this in R  
(preferably with glm() or similar functions)?

(ii) Is there a Bayesian version of the glm()-function where I can  
specify the prior distribution for my regression parameters?

Thanks in advance!
Kind regards,
Mara Pfleiderer