Add captions to onebox images in chat [duplicate]

Question

When I share images with the folks in The Nineteenth Byte, I frequently find myself double posting just so that I can explain what the image is about or why I bothered posting it. For example:

Addind any text other than the link prevents it from being oneboxed. That means that if I have to post the image alone. If someone finds it interesting and stars it, an ugly imgur link gets added to the transcript. Others have to click on it to have a clue what it is about.

My Suggestion

In addition to the current oneboxing...

If the entire message consists of [MESSAGE](LINK) and that link is otherwise oneboxable, just onebox it and display the message above it. For example, the following message would show an image of the weather with "Hello, cold, rainy world!" floating above it.

[Hello, cold, rainy world!](https://i.stack.imgur.com/jYXpI.png)

Answer 1

I really like the idea, but I think the implementation could be improved.

I've seen a couple times where users post images, particularly animated gifs, that they don't want to be flashing around for the hour until enough people talk to make it go away. In those cases, they write exactly your implementation, but with the goal of the image not displaying.

Hello, cold, rainy world!

The idea being that users who care can click the link and see an image. The obvious workaround in this is to not link the whole thing,

Hello, cold, rainy world!

But that's a frustrating workaround, and it will break backwards compatibility.

I don't really have a better idea, but I can imagine this having some cue similar to the "hide" one.

Hide this!

The syntax for that, of course, is >! Hide this!.

I wonder whether something similar could be implemented. For instance, if writing

![Hello, cold, rainy world!](https://i.stack.imgur.com/jYXpI.png)

could do it, that would be less likely to conflict with present habits or backwards compatibility.

This one is more the developer in me talking, and I don't know whether it would be feasible to implement in a wide-spread way, but I can also imagine this.

<!-- Caption: Hello, cold, rainy world -->https://i.stack.imgur.com/jYXpI.png

But again, that's probably too complicated and hard to remember for a majority of users, not to mention there's a lot of overhead involved just in typing it. But at least I can't imagine it would break backwards compatibility.