The question is in the title, really. It is similar to other questions, such as: How much information should I divulge about on-going unpublished research at a conference?
However, I am wondering if publishing an internal (i.e. non refereed, etc.) technical report actually protects me from any plagiarism if I should encounter one.
I know people sometimes publish in arxiv for this purpose, but I have two concerns: a. it is not customary for me to do that, and b. I am not sure it is considered 'unpublished' for the sake of publishing in a refereed journal later.
What do you think?