background images and display:none

[eocene]

I was checking what was causing the "IO_write, closing with pending data
not sent" messages with background images, and at least in the case that
I just dug into, it was due to display:none. Is this considered
expected behaviour, or a bug, or...?

[Johannes Hofmann]

Post by eocene
I was checking what was causing the "IO_write, closing with pending data
not sent" messages with background images, and at least in the case that
I just dug into, it was due to display:none. Is this considered
expected behaviour, or a bug, or...?

The simple case can be prevented with something like in attached
patch. But even with this patch I see quite some of the messages you
mention. So there must be something else going on too.

Cheers,
Johannes

[eocene]

Post by Johannes Hofmann

Post by eocene
I was checking what was causing the "IO_write, closing with pending data
not sent" messages with background images, and at least in the case that
I just dug into, it was due to display:none. Is this considered
expected behaviour, or a bug, or...?

The simple case can be prevented with something like in attached
patch. But even with this patch I see quite some of the messages you
mention. So there must be something else going on too.

It might be that elements with display:none are fixed now but
elements whose ancestors are display:none are still broken.

[Jorge Arellano Cid]

Post by eocene
I was checking what was causing the "IO_write, closing with pending data
not sent" messages with background images, and at least in the case that
I just dug into, it was due to display:none. Is this considered
expected behaviour, or a bug, or...?

An "informative" message.

When the socket has already received the request string and has
asked the kernel to deliver it, but it has not got out fully yet,
and suddenly the socket gets cancelled/aborted, the message is
displayed.

Whether expected behaviour or a bug, depends on the case.