Hi!
Sorry for the delay on this. We had some issues with the fix in QA review that delayed deploying the update. I now expect to deploy it tomorrow morning.
Quickly, you can workaround the issue by adding a rule such as the following to your site-wide css:
iframe.twitter-tweet {
max-width: 315px !important;
}
The explain: The bug is caused because the browser has an undefined resolution for how wide the iframe should be inside of a table-cell, effectively a condition where the embedded Tweet is telling the browser that it has max-width of 100%, and an initial width of 520px and the browsers disagreeing about whether the 520px should expand the size of the table-cell, or if the iframe should be contracted down to the 100% (315px) of the column. To my mind it seems fairly obvious that the iframe should shrink, but the browser behavior is undefined hence this bug renders in Firefox and IE, but not in Safari and Chrome.
We didn’t see this in testing because very few sites still use table-based layouts, and even then (having built numerous test-harnesses for this bug now), the cell-expanding behavior of table-cells is only evoked some of the time.
Previously we fixed the bug by inverting the use of width and maxWidth—the visible result was the same, but the browser’s reasoning meant that the initial width would be taken from the cell width, before constraining the max-width. Unfortunately (and the reason the bug reappeared, I’m afraid) this swap caused a different, much more widespread layout issue in iOS browsers (which has its own share of infuriating iframe behaviour.) The visibility of that issue meant we had to revert this fix in the short-term.
So that’s what’s happened, my apologies for how annoying this is on your sites, unfortunately the intricacy of browser engines has made it hard to fix this robustly. Hopefully the above CSS will sort you out in the short-term, and our fix (which should go out tomorrow) will add special handling for Tweets inside of table cells to prevent them rendering wider than the initial size of the cell.
Thanks for your patience,
Ben
