The first step to access mimi is to set up your School of Computer
Science (SoCS) account on our server called mimi. To do so,
on McGill wifi (or after connecting to the McGill
VPN) visit this page
to set up your SoCS account.
You will be assigned a username (which we will refer as your SoCS username, or short username), which is made of the first letter of your first name, followed by the first five letters of your last name, followed sometimes by a number. You will also be assigned a password.
Save the username and password that you received!
Note that until you setup your ssh key (later), you will need to be on the McGill wifi or VPN to connect to mimi.
The first step to connect to mimi is to open a terminal. On Windows,
you can open the windows menu and search for cmd (or use
winkey+R, then enter cmd). On Mac, you can go to
applications and search for Terminal.
We will now use ssh to connect to mimi. ssh
is a very versatile program that we use to log into a remote computer.
Run the following command:
ssh [user]@mimi.cs.mcgill.caReplacing [user] with your SoCS username. For example, for me this
would be ssh mberub20@mimi.cs.mcgill.ca. This tells the ssh
command that you want to login to the [user] account on the
mimi.cs.mcgill.ca machine.
You will now be prompted for a password. Enter the password you were given earlier. Note that the password will not appear as you type it! This is by design for security. Copy+paste still works, or you can type it manually.
You should now be connected to mimi! You should see
[user]@teach-node-##:~ at the bottom left of your terminal:
the part to the left of the colon ([user]@teach-node-##) indicates that you are
connected to account [user] on machine teach-node-## (which
is one of the nodes of the mimi server). The path to the right
of the colon is your current working directory on this machine:
~ is your user home directory (more on that
soon).
Once connected, you can change your password:
passwdChange to something you will remember. You must use lower/upper case and at least one number/symbol!
Do not log in to Mimi with your first.last McGill account. It might successfully connect you! But you will be logged in as a generic "campus user", not a SoCS student. Many things that you need for this course will not work correctly as campus user.
To logout of mimi and go back to your local terminal, you can type
logout or exit.
You're already used to ordinary passwords, that you type in when you want to log in. But ssh has an alternative mechanism for providing a password. That alternative mechanism is called an ssh key.
By setting up an ssh key on your local device and adding it to mimi, you will be able to connect to mimi from home without needing the McGill Wifi or VPN.
The first step is to generate your local key. From your local computer, open a terminal (do not connect to mimi yet) and type:
ssh-keygenThis will ask to enter a file in which to save the key: you can press enter if you want to use the suggested path in parentheses (I suggest doing this). Note down the path where you saved the key.
This will also ask you if you want to setup a password for the ssh key: you can press enter directly if you do not want a password (remember that the ssh key itself is going to act as your password, so this is just if you want additional protection in case someone that shouldn't have your private ssh key gets access to it).
If you go to the location where the key was saved, you should now see
two files: one private key (which looks like id_ed25519),
and one public key (which looks like id_ed25519.pub).
As the names suggest, you can distribute the public key however you want, but the private key should be kept private! Using ssh keys to replace passwords works as follows: The public key will first be installed on the machine you want to connect to (in this case, mimi). When you want to connect to that machine at a later time, it will use that public key to craft a challenge that can only be solved by the person who has the private key in their possession. This then proves that you have the authority to log in.
The first step will thus be to install this public key on mimi! To do
this, we'll need to copy the public key from the local computer to mimi.
The command to do this is scp, which is already installed
on Mac, Windows and Linux out-of-the box. Run the following line on your
terminal:
scp [path to the public key] [user]@mimi.cs.mcgill.ca:For example, on Windows this could look something like
scp C:\Users\mathi\.ssh\id_ed25519.pub mberub20@mimi.cs.mcgill.ca:
scp takes as input two arguments: the location of the file you want
to copy, and where you want to copy it to. Since the location you are
copying it to is remote, the address is composed of
[user]@mimi.cs.mcgill.ca, which is the machine and user
you're sending it to, a colon :, and the path where you
want to send it on this machine. Here, since we didn't specify a path
after the colon, it will send it to our home directory
(~) by default.
Now, connect to mimi. If you do ls to list the files in
the current directory (which is your home (~) directory by
default), you should now see your id_ed25519.pub file
there. The final step will be to include this file in
~/.ssh/authorized_keys.
To do this, create a .ssh directory by typing mkdir .ssh
from your home directory (the current path after
[user]@mimi.cs.mcgill.ca: should be ~). Since
this directory starts with a dot, it is a hidden directory: If
you type ls, you will still not see it, but if you type
ls -a, where the -a flag tells the
command to show all files and directories, you should now see
.ssh.
When logging in, ssh looks at the public keys inside the
file ~/.ssh/authorized_keys. This file is simply a
concatenation of public keys, as such, everytime you want to add a new
key you will simply add it to the end of this file.
If you have never setup a ssh key on mimi yet, this file doesn't
exist yet, so we can simply move the public key file to .ssh and rename
it authorized_keys. The mv command can do both
of this at the same time: -
mv id_ed25519.pub .ssh/authorized_keys
However, if you have already set up a ssh key in the past, the
previous command would overwrite it with the new key! If you already
have keys (for example if you want to access mimi from two different
computers), what you want to do instead is to append the key to
this file by doing the following: -
cat id_ed25519.pub >> .ssh/authorized_keys
We will talk more about the >> operator in a
couple of weeks.
That's it! You should now be able to connect to mimi from your local device without having to be on the McGill wifi or VPN! Let us know on Ed if you had any issues!
Assignments this semester will be done on git and autograded. To get these assignments, git needs to confirm your identity: just as in the previous parts, this will be done with ssh key! There are two differences this time:
Download the ssh key you received by email to your computer. Assuming
your name is Jane Smith, the file you downloaded should be named
something like jsmith13-key.
You will next need to transfer this key into your SoCS account.
Transfer the file to your mimi home directory using scp. Try to do this by yourself, now that you've already done this once in the previous part!
Next, log into mimi and do the following:
chmod 400 jsmith13-key
mv jsmith13-key .ssh.ssh directory config inside the
.ssh directory.Congratulations, you set up your ssh key. You will now be able to obtain and submit the homework via git on mimi.
comp-206 in your home
directory.homework, labs, and
notes as subdirectories of comp-206.homework directory and clone the repository
containing these instructions:
git clone fall2026-comp206:hw1.githw1.Check you did it right: ls ~/comp-206/homework/hw1/.git
should succeed and show a listing of some strange, internal git
files.
hw1 directory, create a file called
hello_world.txt using the touch commandNext:
tree -F comp-206.
This will output a textual diagram of the directory structure you
created in part 1.tree -F), clearly showing your username.hw1 directory, and rename it to screenshot.png
(or screenshot.jpg if it happened to be in jpg
format.)To recap, your hw1 directory should contain at this point:
hello_world.txt; andscreenshot.png (or screenshot.jpg);
andinstructions.md (this file!)Then,
git add . to prepare all changes in the repo to be
committed.git commit -m "completed assignment 1"git push to upload your solution, which will run
the autograder on it. The autograder will output a message saying
SCORE telling you whether you passed the assignment or
not.If you need to revise your solution and resubmit it, repeat the three
steps in this section: git add .,
git commit -m "a message describing the changes",
git push.
You can run the autograder via git push as many times as you like!
(In practice when doing software development, running
git push won't run an autograder, but it will typically run
some scripts to run a test suite on your code, compile your code, and
deploy your program, i.e. make it accessible to users!)
This assignment is graded simply on completion, and the main point of it is to familiarize you with the mechanism by which you'll be doing and submitting your homework throughout the semester.
Simply pushing with the required files present and of the correct types will earn full marks.