Publishing from a terminal or from a browser

Both take a folder and return an address. They differ in what you have to have open.

Both take a folder of finished files and return a working address. The difference is entirely about who is publishing.

A one-line publish command beside a browser upload panel, both returning an address.
A one-line publish command beside a browser upload panel, both returning an address.

Command line publishing

One command in the folder, and the page is live in a few seconds.

Its real advantage is not speed, it is that a command can go into a script. Build and publish become one step, and the publish stops being something a person remembers to do.

For a developer already in a terminal this is the lowest-friction route that exists, and nothing with a browser interface beats it.

Browser publishing

Open a page, drop a folder, copy the address.

Nothing installed, nothing to remember, and no terminal. That matters because the person who maintains a page is often not the person who built it, and requiring a terminal means every update goes through somebody who has one.

Pages maintained by a queue go stale. That is the real cost and it does not appear in any comparison table.

Command line Browser
Install needed Yes No
Scriptable Yes No
Who can publish Developers Anyone
Part of a build Naturally No
Correcting a typo Re-run Re-upload or edit

Reliability is not the difference

Both are putting static files on a server, and static files are the most reliable thing on the internet.

Whatever uptime difference exists comes from the hosting arrangement behind each service, not from whether the upload came through a terminal or a browser. It is not a useful axis for choosing.

The same page with two maintainers, one in a terminal and one who is not.
The same page with two maintainers, one in a terminal and one who is not.

The correction question

A price is wrong. One line.

Both routes default to re-publishing the whole folder, which means having the folder, remembering which version is current, and repeating the upload. For a developer, trivial. For anyone else, an obstacle, and obstacles are why pages sit wrong for months.

If the page changes often, prefer whichever option lets you edit the published page directly. That is the axis that matters more than either interface.

Using both

There is no conflict.

Project pages that rebuild from source go through the command line, in the build script, automatically. One-off documents, a brochure, a report, a price list, go through the browser, published by whoever wrote them.

Choosing one tool for both is how you end up with either a developer uploading brochures or a marketer being asked to install a terminal.

Two neighbouring cases are worth a look: Comparing the quick ways to publish a page and Choosing between a document host and a build platform.

Put it at an address

Match the route to whoever publishes, script it when it repeats, confirm the address is permanent, check how a typo gets fixed, and never make the page's maintainer learn a terminal to correct a line.

Questions people ask

What is command line publishing good for?

Speed and repetition. One command publishes, and it goes into a build script so publishing stops being a manual step.

What is the browser route good for?

Everybody else. No install, no terminal, and the person who wrote the page can publish it without asking anybody.

Is one more reliable?

No. Both are serving static files. Reliability comes from the hosting behind them, not from how the files arrived.

Which handles corrections better?

The one that lets you edit the page directly. Both re-upload by default, and re-uploading is why stale pages stay stale.

Can I use both?

Yes, and many do. Command line for project pages, browser for one-off documents.

Keep reading