1. box-pack, box-align, and box-direction are insufficiently direction-agnostic in their descriptions. While they do say something about what direction or edge they are referring to based on the box-orient value, the actual algorithm descriptions tend to assume a horizontal box-orient. These need to be reworded slightly to be properly agnostic about the direction (or conversely, to be properly *explicit* about the direction, whichever is clearer). 2. `box-align: stretch` does not allow sufficient control over the alignment of the box contents after the height is increased. I guess the children just act like normal boxes, and top-align themselves? This prevents us from, say, baseline-aligning boxes but still wanting a background/border to be the same height on everything. I suggest that either 'stretch' should be a valid value *in addition to* the other values, or that we should define vertical-align affecting flexbox children similarly to how it affects table cells. ~TJReceived on Tuesday, 13 April 2010 18:53:05 UTC
This archive was generated by hypermail 2.4.0 : Monday, 23 January 2023 02:13:45 UTC