IFRAME sizing regression

Commits 1d4a172ec270987c8785651dcc7a563b3001095f and 3c713dd

Tue, 18 Feb 2014 10:05:52 +0000

Commits 1d4a172ec270987c8785651dcc7a563b3001095f and 3c713dd3a2e31c8a15517d1d50c2b8d6aad853ea break IFRAME sizing of several HTML newsletters while I didn't experience any issues before those patches. FF 27 Linux.

Example 1, works with the 2nd commit reverted.

Tue, 18 Feb 2014 10:07:52 +0000

Example 1, works with the 2nd commit reverted.

Example 2, works with both commits reverted.

Tue, 18 Feb 2014 10:08:46 +0000

Example 2, works with both commits reverted.

> Commits 1d4a172ec270987c8785651dcc7a563b3001095f and > 3

Tue, 18 Feb 2014 17:42:43 +0000

> Commits 1d4a172ec270987c8785651dcc7a563b3001095f and > 3c713dd3a2e31c8a15517d1d50c2b8d6aad853ea break IFRAME sizing of > several HTML newsletters while I didn't experience any issues before > those patches. FF 27 Linux. Not sure what to do. "Height: auto" absolutely cannot be used, since it breaks both dynamic and smartmobile views (it causes automatic scrolling of the containing div, which breaks things like automatic hiding of the addressbar in smartmobile, which is 100% unacceptable). So these are not revertable Not to mention that I have a 150 message mailbox of test messages that at one point did not display correctly. The current code does as good of a job as before, without breaking anything (no solution I've ever come up with has displayed 100% of messages correctly). If someone else wants to try and come up with a better solution, that would be great. I've spent hours on this with minimal luck, and there really is very little documentation I've found on the web that can help. Two things need to be avoided: 1. Resizing based on defer/delay code 2. Height: auto

Changes have been made in Git (master): commit 718dc9394dc9

Fri, 07 Mar 2014 08:59:46 +0000

Changes have been made in Git (master): commit 718dc9394dc9639bf7093461ee2c0bb8f8e1c23c Author: Michael M Slusarz Date: Fri Mar 7 01:26:16 2014 -0700 Bug #12978: This seems to work at least slightly better The simple fact is that computed body height may not be the same as computed html tag height. Even stranger, the latter is oftentimes greater than the former, even though it is embedded within the former. imp/js/imp.js | 9 ++++++--- 1 files changed, 6 insertions(+), 3 deletions(-) http://git.horde.org/horde-git/-/commit/718dc9394dc9639bf7093461ee2c0bb8f8e1c23c

Changes have been made in Git (master): commit d50926e82b7e

Tue, 18 Mar 2014 06:11:16 +0000

Changes have been made in Git (master): commit d50926e82b7e48557373963f0668a91f31e1c833 Author: Michael M Slusarz Date: Tue Mar 18 00:11:02 2014 -0600 Bug #12978: More tweaking of IFRAME sizing algorithm imp/js/imp.js | 19 +++++++++++++------ 1 files changed, 13 insertions(+), 6 deletions(-) http://git.horde.org/horde-git/-/commit/d50926e82b7e48557373963f0668a91f31e1c833

Marking as resolved. This is always going to be a WIP with

Tue, 18 Mar 2014 06:12:08 +0000

Marking as resolved. This is always going to be a WIP with constant tweaking.

This has much improved now. The only case where it still doe

Tue, 18 Mar 2014 09:19:48 +0000

This has much improved now. The only case where it still doesn't work correctly are HTML messages with images that don't specify sizes, so that the body resizes while the images load. Going back to that message later when all images are cached works perfectly. Any chance to trigger another resize just in case the body size changed when the document has completely loaded? This would reduce the flickering to messages that don't specify image sizes.

> Any chance > to trigger another resize just in case the

Tue, 18 Mar 2014 18:23:45 +0000

> Any chance > to trigger another resize just in case the body size changed when the > document has completely loaded? This would reduce the flickering to > messages that don't specify image sizes. We already do this. In fact, the iframe resize code is triggered after each individual image is loaded. The issue might be we need to add a delay to the callback to the resize method after the image onload event is triggered, because it is possible the onload is triggered before the image is inserted/sized in the IFRAME DOM element.

Changes have been made in Git (master): commit c5743be34d4d

Tue, 18 Mar 2014 19:15:19 +0000

Changes have been made in Git (master): commit c5743be34d4d9ef90b558c645cf09dd594850e51 Author: Michael M Slusarz Date: Tue Mar 18 13:12:14 2014 -0600 Bug #12978: Fix resize behavior when images are not blocked imp/js/imp.js | 28 +++++++++++++++++----------- 1 files changed, 17 insertions(+), 11 deletions(-) http://git.horde.org/horde-git/-/commit/c5743be34d4d9ef90b558c645cf09dd594850e51