Adding iframe-Resize to disconnected iframe fails, even after it is add to the page.

[maximedasilva]

Describe the bug
In no logs mode (log: -1), some logs are still presents, even if they're useful to debug, they should be absent when the configuration explicitely sets logging to be disabled.

To Reproduce

1. Set up iframe-resizer with the log: -1 option.
2. Introduce an error within the setup
3. Observe the browser console.

I set up a minimal repro https://codesandbox.io/p/devbox/prdpzk which is dumb but explains the point

Expected behavior
The log: -1 option should globally and completely suppress all console output and internal logging related to the library

Screenshots

Desktop (please complete the following information):

• OS: Mac OS 26.0.1
• Browser: chrome (v 141.0.7390.108)
• Version: 5.5.7

Additional context
this already has been discussed here :
#1526

[davidjbradshaw]

Hi,

Your ticket raises two very different questions for me.

• The first is a philosophical one about whether warnings and errors should be disableable. The logging is controlled by the separate auto-console-log project and this is really a question for over there.

• The second is should the edge case you have come up with of running iframeResize() on a disconnected iframe cause this warning to be displayed in the first place and I think properly not. So I will have to give it some thought about how to detect if an iframe is not actually currently connected into the DOM (iframe.parent = undefined?) and suppress this warning in that case. That said I'm a bit confused as to why you get the warning a second time once the iframe has been added into the page. I'm going to have to run your test case locally and investigate this a bit more.

[davidjbradshaw]

Yep, looking at this a bit more. If you attach iframeResize() to a disconnected iframe, it will fail to send messages to the iframe, even after it has been added into the page.

[maximedasilva]

Yes, you are right, my example is more complex than I thought. It seems we have an issue here.

But my point was, besides this bug, I'm not sure – with my comprehension of your work – that the philosophy behind the warnings and errors logs has to be discussed in the auto-console-log repo. Correct me if I'm wrong, but I believe that this is handled by this part of code, but I might be wrong.

Isn't the outputSwitch function responsible for this ?

[davidjbradshaw]

I might have been mistaken about which project selects the logging level.

That said, I still feel that an error should always be displayed when there is an internal error in iframe-resizer that prevents it from executing correctly. Their are very few of these and every time someone has raised this as an issue, it has uncovered an obscure bug that would not otherwise have been fixed.