Written 30 Aug 2026. This is a how-to, not a product card.

Download a file onto the computer

After SSH works, pull one file from a web address with curl so the computer has the file without a USB stick. You don't need a browser on it for this.

Wrap long lines kept long lines readable in the window. Tonight you pull one file from a web address onto the computer. curl does that. You pick a save path in home and a URL. It fetches the file. You are still only practicing. It does not need a desktop browser, a USB stick, or a new mini PC for this one job.

Skip until SSH already works

SSH from the couch should already get you a prompt. If the computer cannot reach past the house router yet, finish check that the computer can reach the internet first. If you only need a file from your laptop, that is copy a file to the computer from your laptop. If you only need long lines wrapped, that is wrap long lines. If you only need line numbers, that is number lines. Skip this whole note if you only need a house share other computers browse. That is a folder the house can open.

What this is

curl means pull bytes from a web address. You type curl, a save flag, a path in home, then a URL that starts with https://. -o means save as this name. -O is optional keep the name from the URL. -L is optional follow a redirect. -I is optional headers only, no file body. It writes the file in home. It does not open a browser. It is not a copy from your laptop. That is copy a file to the computer from your laptop. It is not a ping past the router. That is check that the computer can reach the internet. It is not wrapping long lines. That is wrap long lines. It is not piping a download into a shell. Skip that tonight.

Download one practice file into home

SSH in. Make sure home is where you sit:

pwd

You want a path that ends in your login name. Now pull one small practice page:

curl -o ~/example.html https://example.com

-o ~/example.html is the save path. https://example.com is a boring public page used for practice. You want a short progress line, then a quiet prompt again. Peek:

ls -lh ~/example.html

You want a small file size, not zero bytes. Short dump:

head -n 5 ~/example.html

You want HTML words about Example Domain. That is the whole first-night trick: one file from a URL, saved in home, no USB stick.

Need the filename from the URL instead of picking one?

curl -O https://example.com/index.html

-O keeps index.html as the name in the folder you are sitting in. Prefer -o ~/name tonight so you choose the path.

Need to follow a redirect when a site moves the file?

curl -L -o ~/example.html https://example.com

-L is optional. Plain curl -o is enough when the URL already lands on the file.

Useful extras, still looking only

Need headers only, no body saved?

curl -I https://example.com

You want lines like HTTP status and content type. That check does not rewrite home.

Skip curl … | bash tonight. A download into a shell is how people run installers they have not read. Tonight you only save a file you can open with head or less.

Skip wget as tonight's lesson. Some computers have it. Tonight's tool is curl.

Skip -k (skip certificate checks) and huge OS images tonight. Practice one small page first.

If it says command not found for curl, install it, then try again:

sudo apt install curl

Debian and Ubuntu Server usually already have it. That install is not a reason to buy a browser for the computer.

When you are done looking, throw away the practice file:

rm ~/example.html

Careful delete only. That is delete a file carefully. If you also made index.html with -O, delete that one the same way.

What this is not

This is not a copy from your laptop. That is copy a file to the computer from your laptop. It is not a ping past the house router. That is check that the computer can reach the internet. It is not wrapping long lines. That is wrap long lines. It is not line numbers. That is number lines. It is not a short dump of a file you already have. That is read a file without opening the editor. It is not one screen at a time. That is read a file one screen at a time. It is not first lines. That is see the start of a file. It is not a quiet save of command print. That is save output to a file. It is not a house share other computers browse. That is a folder the house can open. It is not opening SSH to the internet so you can download from a coffee shop. That is keep the computer on the house network.

What next

Need to check the file fingerprint after the download? Check a file fingerprint. Need a file from your laptop instead? Copy a file to the computer from your laptop. Need to prove the computer can reach past the router first? Check that the computer can reach the internet. Need a short dump of the file you just pulled? Read a file without opening the editor. Need one screen at a time? Read a file one screen at a time. Need the throwaway gone for real? Delete a file carefully. Keep SSH on the house network: keep the computer on the house network. Don't forward port 22.

Skip

• Skip if SSH still fails. Finish SSH from the couch first.
• Skip if internet checks still fail. Finish check that the computer can reach the internet first.
• Skip curl … | bash, random installer URLs, and files you have not peeked at. Practice save with -o, then peek with head or less.
• Skip -k, huge OS images, FTP tonight, and uploads with -T. One small download is enough.
• Skip saving into /etc, SSH folders, or paths you don't own. Practice only in home.
• Skip opening SSH to the internet so you can download files from a coffee shop. That is keep the computer on the house network. Don't forward port 22.