I know svn can handle this, for example the svn repo would have directroies such as
svnrepo/trunk/
pc1
pc2
...
pcn
then on user only needs to checkout svnrepo/trunk/pc1, and another checkouts trunk/pcn.
But if I use git, until now, I know all users will have all the files, for example, there will be pc1, pc2...pcn on while only pc is needed
Sometimes this could be confusing.
So is that possible to have one git repo but each dir is distributed indepdendently from the others?
In my honest view you’re trying to use git the same way you use svn.
Using git you won’t accomplish this (and even if you can using some strangely advanced commands) I would never recommend that.
What you can quite easily do is have a different .gitignore in each pc and create a process to update it without merging and resolving conflicts which would always exist. Of course this is definitely an expensive an not necessary process.
The concept in git is everybody have the whole repository. This is not a drawback. SVN probably have such mentioned capability due to the network traffic it uses and the lack of a decent merging strategy.
Whatever you are trying to do, I suggest you trying in the “git” way, which means lots of commits and lots of merges without any fear.