- From: Javier Fernandez <jfernandez@igalia.com>
- Date: Wed, 29 Jul 2015 11:04:00 +0200
- To: fantasai <fantasai.lists@inkedblade.net>, www-style list <www-style@w3.org>
Hi fantasai, Thanks for the replies, really helpful. On 07/28/2015 05:25 PM, fantasai wrote: > On 07/27/2015 05:12 PM, Javier Fernandez wrote: >> * flex-start: >> - Aligns the alignment subject to be flush with the edge of the >> alignment container corresponding to the flex container’s main-start or >> cross-start side, as appropriate. > > This will always be the cross-start side, since align-self operates in > the cross dimension. > So, this is an error in the specs, isn't it ? I was about to ask what would make we use either main-start or cross-start side, but from your words I assume it's a mistake and we should use only cross-start side. >> Since align-self applies along the cross axis, I assume that container’s >> start edge, being the container the flex line, will be the same than the >> cross-start side, independently of the flex-flow value or the >> writing-mode. I understand that both, start and flex-start produce >> always the same results. am I right on this assumption ? > > No, not correct. For align-self, > * flex-start/flex-end depend on flex-wrap of the container > * start/end depend on writing-mode of the container > * self-start/self-end depend on writing-mode of the item > > If flex-wrap is wrap-reverse, then the start and flex-end alignments > will be identical, and start/flex-start will be opposite. > Correct, I understand now the difference. I totally forgot about flex-wrap. >> Another doubt I have is how to deal with self-start and orthogonal >> flows. I think it d be something similar to what happens in Grid Layout, >> hence orthogonal flows lead to self-start equivalent to start and >> flex-start. Only when using parallel flows, but opposite directions the >> result of self-start would imply any difference. > > No, when determining start/end (or self-start/self-end) you always use > the writing mode. It may vary whether you are mapping using the block > axis directions or the inline axis directions, but you are mapping > against the writing mode in one of those directions. > >From the example above, I understand that you mention writing-mode to refer both properties, -webkit-writing-mode and direction. That's make sense. > For example, if we have a vertical-rl ltr item, its start sides are > the top and the right, and the end sides are the bottom and left. > If the cross axis is vertical, the align-item: self-start aligns to > the right. I guess here is "top", isn't it ? Since cross axis is vertical we align flex item top-bottom, If I understood it correctly. If the cross axis is horizontal, self-start aligns to the > top. > Hence here it should be right. > Does that make sense? I hope the example had the mistakes I mentioned, which it'd mean it makes sense for me, otherwise I think I'll have to read it again more carefully. BR -- javi
Received on Wednesday, 29 July 2015 09:05:07 UTC