Is there a reason you shouldn't mix ssh and smart http with git?

I installed a git server using gitolit for a small team. One of us created a repo and pushed the initial commits using smart http - suppose it is on

http://ourhost/project.git

Then I went on to clone the project using ssh ie

git clone git@ourhost:project.git

Git Let me do this and created a project folder, but reported that the repo was empty . Also, any git pull I did gave me " could not find remote ref errors ". When i started

git remote -v show origin 

He showed the HEAD branch: (unknown). When I checked on the server (starts gitlab), I could see the main branch and all the code so that it was clearly there.

, smart http lo, , .

? , , /? Gitlab URL ( ssh smart http) , , ...)

- ...?