IFI-8410 Course Tools

Command Line (CLI) tools for manage class content and submissions.

A guide for students. No prior experience with Git or the command line is assumed. Everything here is typed into a terminal on the course server.

You have four commands. They all start with the course number, so you never have to remember which course you are in:

Add --help to any of them to see its options.

1. Getting a terminal

A terminal is a window where you type commands. There are two ways to get one, and they are equally good — the commands are identical in both.

In JupyterLab

1. Open JupyterLab in your browser and log in.
2. File → New → Terminal (or click the Terminal tile in the Launcher).
3. A black window opens with a prompt that ends in $. Type there.

Over SSH

From the Terminal app on a Mac, or Windows Terminal / PowerShell on Windows:

ssh yourid@<the server name your instructor gave you>

Type your password when asked. Nothing appears while you type it — that is normal.

A notebook cell is not a terminal. You can run these commands in a notebook with %%sh, and section 9 shows how, but a notebook cannot move you into a folder or turn on your Python environment. Use a terminal for those.

2. Your first command, every day

ifi8410-status

Run it the moment you sit down. It looks at everything, prints one line per check, and then tells you what to do next. It never changes your files without asking.

Course: IFI8410 - Programming for Business

⚠️  Course directory       you are in ~
✅  SSH key                ~/.ssh/id_ed25519
❌  Conda environment      none active (expected conda-python3.12)

What to do next
  The conda-python3.12 environment is not active, so your Python packages are missing.
      source conda-env conda-python3.12

Your project is not in this directory. You can work in:

  1) go to ~/IFI8410FALL2026/ifi8410-abc123 and activate conda-python3.12
  2) go to ~/IFI8410FALL2026/ifi8410-abc123
  3) activate conda-python3.12 here, without moving

Which one? [1-3, or Enter to stay here]

Type 1 and press Enter. You are now in your project folder with Python ready, and you can start work. Press Enter on its own and nothing happens.

The one-key version

ifi8410-status --auto

or ifi8410-status -a for short. It does the same three things — takes you to your project, puts you on the work branch, turns on the Python environment — without asking any questions. This is the fastest way to start.

3. Reading the report

The nine rows, in the order they appear:

Under the rows is What to do next — one line per thing that needs attention, with the exact command underneath it. You can always copy that command and run it.

4. The routine

ifi8410-status -a      →  start the day
ifi8410-update         →  get the new material
   ... do your work, in JupyterLab or an editor ...
ifi8410-test           →  save it and run the tests
   ... fix what the tests found, run ifi8410-test again ...
ifi8410-submit         →  hand it in

You can run any of them as many times as you like. None of them ever throws work away.

5. ifi8410-update — get your instructor’s new files

ifi8410-update

Run it at the start of every session and whenever ifi8410-status shows ⚡ Instructor updates.

Updated.
New files from your instructor are now in your repository.
Now at: 8f2c1a4 Week 3 notebook and data

Two things it needs first, and it will tell you if they are missing:

• you must be on the work branch (ifi8410-status offers to put you there);
• your own changes must be saved into Git first — run ifi8410-test, which saves them.

When you and your instructor changed the same file

This is normal and nothing is lost. You will see:

You and your instructor have both changed things since your last update.

Nothing is lost either way: where you both changed the same file, one
version keeps the file name and the other is saved next to it.

  1) Keep my instructor's version; save mine beside it   (recommended)
  2) Keep my version; save my instructor's beside it
  3) Do nothing for now

Which one? [1]

Press Enter to take the recommendation. Both versions of every file survive — one keeps its name, the other is saved next to it with a longer name, so you can open them side by side and copy across anything you need.

6. ifi8410-test — save your work and run the tests

ifi8410-test

It shows you what it is about to save, asks once, then saves it, sends it to GitLab, and starts the automatic tests.

These changes will be committed:
  - unstaged: HW01/exercise1.py
  - untracked: HW01/notes.md

Commit these changes and start testing? [y/N] y
Committed: 3d9f0aa course-test: 2026-09-07 10:15

Testing started.

Tested commit: 3d9f0aa course-test: 2026-09-07 10:15
Pushed:        work
Merged into:   testing

The pipeline is running now on testing.
A failing test changes nothing: fix the problem and run course-test again.

Add a note about what you did with -m:

ifi8410-test -m "finished exercise 1"

The tests run on the server and take a little while. You can keep working while they do. A failing test is information, not a punishment — fix it and run ifi8410-test again, as often as you like.

7. ifi8410-submit — hand work in for grading

ifi8410-submit

This marks the version you want graded.

You are about to submit:
Course: IFI8410 - Programming for Business
Branch: work  ->  grading
Current time: 2026-09-07 10:22 EDT

Submission recorded.

Submitted commit: 3d9f0aa finished exercise 1
Recorded on:      grading
Time:             2026-09-07 10:22 EDT

This is the version that will be graded.

Things worth knowing:

• Submit early, submit often. You can run it as many times as you like before the deadline. The newest submission is the one that counts.
• Nothing is deleted. Every submission stays in the history; a new one just moves the marker.
• ifi8410-status always shows you what is currently up for grading: ✅ Submitted for grading this commit (3d9f0aa) means the version on the server is exactly what you have now. ⚡ ... 4 newer commits not submitted means you have done work since you last submitted — run ifi8410-submit again if you want it graded.
• Name the assignment if your instructor asks you to: ifi8410-submit --assignment HW01

8. Feedback

When your work has been marked, a new row appears:

✅  Feedback               graded-HW01-Sept-16-2026

and below the report:

Read your feedback with: git show origin/graded-HW01-Sept-16-2026 --stat

Copy that line and run it. Your own work is untouched by feedback.

9. The Python environment

Your course packages live in an environment called conda-python3.12. To turn it on:

source conda-env conda-python3.12

To have it on automatically in every terminal from now on — do this once:

source conda-env --set-profile conda-python3.12

Type source conda-env ..., not conda activate .... On this server conda on its own is not set up in a fresh terminal; the conda-env helper does the setup for you. The word source matters: it lets the command change your terminal rather than a copy of it.

10. Using the commands in a notebook

You can run them from a notebook cell with %%sh, and this works well for looking:

%%sh
ifi8410-status --auto

--auto is the right form in a notebook: it never reports an error to Python, so your cell will not turn red. For just the table of rows:

%%sh
ifi8410-status --quiet

Two limits, and they are not bugs:

• A cell cannot move you into a folder or turn on the environment. It runs in a throwaway shell that disappears when the cell finishes. The commands know this: instead of a menu they print the two lines for you to run in a terminal.

• A cell cannot ask you a question, so no menus appear there. ifi8410-test and ifi8410-submit from a cell need --yes to answer their one question in advance:

  %%sh
  ifi8410-test --yes -m "week 3 exercises"
  

If a plain %%sh ifi8410-status cell ends in a red CalledProcessError, that only means a ❌ row was found — the report above the traceback is the real message. Use --auto, or start the cell with %%sh --no-raise-error.

11. When something goes wrong

Work down the report from the top: the first ❌ is usually the cause of everything under it.

❌ SSH key — no ~/.ssh/id_ed25519

You have no ID card for GitLab yet. Run the line the report gives you:

mkdir -p ~/.ssh && chmod 0700 ~/.ssh && ssh-keygen -t ed25519 -N '' -C 'yourid@ifi8410' -f ~/.ssh/id_ed25519

Then show the public half and copy the whole line it prints:

cat ~/.ssh/id_ed25519.pub

Paste it into GitLab at https://git.insight.gsu.edu/-/user_settings/ssh_keys (Add new key → paste → Add key). Run ifi8410-status again.

❌ GitLab connection — cannot reach ...

Read the sentence under it; the advice depends on the cause. Most often you are off the university network or VPN — connect and try again. If it mentions the host key, run the ssh-keyscan line it prints; that is a one-time step.

❌ Branch — main (your work belongs on work)

Answer y when it offers to switch, or use ifi8410-status --auto. Your work belongs on work; the other branches are managed for you.

“Your project is not in this directory”

You are somewhere else in the file system. Pick a number from the menu, or run ifi8410-status --auto.

“You have more than one ifi8410 project on this computer”

You have cloned more than one project. Pick the one you mean from the list, or cd into it first. --auto will not guess for you.

“Your work is not ready to update”

ifi8410-update needs your changes saved into Git first. Run ifi8410-test, which saves them, then update.

You answered the menu but your terminal did not move

Your terminal session started before the tools were installed or updated. Fix it for this session:

source /etc/profile.d/course-tools.sh

Every new login does this for you automatically.

12. Quick reference

ifi8410-status              check everything, then offer to fix where you stand
ifi8410-status --auto       go to the project, switch to work, activate Python
ifi8410-status --quiet      just the table of rows
ifi8410-status --offline    skip the checks that need GitLab
ifi8410-status --path       print the folder your project is in

ifi8410-update              bring in new instructor files
ifi8410-test                save your work and run the tests
ifi8410-test -m "note"      ... with a note about what you did
ifi8410-submit              hand in the version to be graded
ifi8410-submit --assignment HW01

source conda-env conda-python3.12               turn Python on now
source conda-env --set-profile conda-python3.12 turn it on in every terminal

Useful facts:

• Your course folder is ~/IFI8410FALL2026.
• Your project is inside it, named by GitLab — something like ifi8410-abc123.
• Your work belongs on the branch called work.
• ~ means your home folder; cd ~/IFI8410FALL2026/ifi8410-abc123 moves you into your project by hand.

13. Getting help

Run ifi8410-status and copy the whole report, including the lines under What to do next, into your message to course staff. That report tells us in one glance what your set-up looks like, and it is much faster than a description.

None of these commands deletes your work, rewrites your history, or discards a file you have edited. If something looks alarming, stop and ask — nothing is lost.

14. Working on Your Personal Computer

You can clone your Git repository on your personal computer. However, the course tools are not available and > your must use the git command.

You must keep each cloned version updated!

If you maintain multiple cloned copies on different systems, you must keep them in sync with your Git repository in order to receive new updates.

On your laptop, use a plain git pull, after first committing your own changes. The release lands on the work branch on GitLab, so pulling that branch is all it takes:

git switch work                      # the branch releases go to
git add -A
git commit -m "my work"              # commit first, or the pull can refuse to run
git pull --no-rebase origin work
git push origin work                 # so GitLab and your cluster clone get your work too

• --no-rebase makes the pull a merge, which is what ifi8410-update does. Some laptops are set up to rebase on pull by default, and a rebase rewrites your commits.
• Check it worked: ls .course/release-notices/ should show the latest version
• If Git reports a conflict: you edited a file that the release also changed. You can fix the files and commit with git add -A && git commit, or back out with git merge --abort. Notebook conflicts are painful to fix by hand. The easier route is to abort, push your work, and run ifi8410-update on the cluster, which handles that case.
• Before switching machines: always push from one before pulling on the other. Otherwise the same file can change in both places and conflict.