Every design blog will tell you drag-and-drop is the modern way to accept input. I tried it, thought about it properly, and then shipped the most boring thing possible: a plain text input. Here is why that was the right call for a video downloader.
The input is a link, not a file
A drag-and-drop zone is built for files. You drag a photo from your desktop, the browser fires a drop event, and event.dataTransfer.files gives you the file. That is a lovely flow — for an image editor, a file converter, a cloud drive.
My site is not any of those. The thing a visitor brings is a URL, copied from TikTok, Instagram or Facebook. A URL is not a file. There is nothing on their desktop to drag. The only way a link gets onto the page is through the clipboard: copy, switch tab, paste.
You can drag a link. Dragging a URL from the address bar into a drop zone fires a drop event with text/uri-list data instead of files. It works in Chrome and Firefox, mostly. But it is a party trick, not a workflow. Nobody copies a reel link by dragging the address bar. They long-press, tap copy, and switch apps. On a phone — where most of my visitors are — dragging a URL is not even possible.
The plain input does more work than it looks
A single <input type="url"> with the right attributes is doing a surprising amount:
- Mobile keyboards know what to do.
type="url"brings up a keyboard with.and/on the first screen, and no autocapitalisation mangling the link.autocapitalize="off"andautocorrect="off"stop the phone helpfully "fixing" a URL that was already correct. - Paste is the real feature. The
pasteevent fires the moment a link lands in the box. I read the clipboard text from the event, trim it, and the resolve can start without the visitor tapping anything else. No drop handler, no dragover styling, no "release to upload" state to maintain. - Keyboard-only visitors are not locked out. A drop zone is a mouse gesture. An input works with Tab, Ctrl+V and Enter. I keep the focus outline visible and the label tied to the input with a real
<label>, so a screen reader announces exactly what the box is for. - It is one line of HTML, not a state machine. Drag-and-drop needs dragover, dragleave, drop, and a fallback for the browsers that handle
text/uri-listdifferently. Each of those states is a chance for a bug a visitor can see. The input has no states. It is either empty or it has a link in it.
What I gave up, and what I kept
I did give up one thing: the visual drama. A big dashed box with an icon looks impressive in a screenshot. A plain input looks like nothing at all. But the homepage of ramrom.io is a tool, not a portfolio piece. The visitor is not there to admire the input; they are there to paste a link and get a video. Every extra visual element is a moment of attention spent on the container instead of the job.
What I kept was the affordance that actually matters: the box is wide, it is the first thing on the page, and the placeholder text says what goes in it in plain words. A visitor who has never used the site knows what to do in under a second, on any device, with any input method.
The rule I took from it
Fancy input patterns are worth it when they match what the visitor is actually carrying. A file converter should accept dropped files. A downloader should accept pasted text. When the pattern and the payload do not match, the fancy version is not an upgrade — it is a demo of a feature nobody uses, maintained at the cost of the one everybody does.
Build the input for the thing people bring, not the thing that looks good in a screenshot.
