W3C home > Mailing lists > Public > www-style@w3.org > November 2015

Re: [css-text] pre-wrap / pre-wrap-auto

From: Florian Rivoal <florian@rivoal.net>
Date: Mon, 9 Nov 2015 16:30:14 +0900
Cc: "www-style@w3.org" <www-style@w3.org>
Message-Id: <701BEACF-B991-459F-B841-3EDB8DE6BFE1@rivoal.net>
To: fantasai <fantasai.lists@inkedblade.net>

> On 09 Nov 2015, at 13:55, fantasai <fantasai.lists@inkedblade.net> wrote:
> Proposal:
>  1. Remove the pre-wrap-auto value that we added to css3-text 5 months ago.
>  2. Make pre-wrap UA-defined such that all of the following behaviors are allowed:
>      a) wrapping after entire sequence of spaces
>      b) wrapping after each space
>      c) "visually collapsing" spaces that would otherwise need to be wrapped
>      d) hanging trailing spaces (as in 'hanging-punctuation')
>      e) anything else we think should be allowed
>  3. Possibly recommend some specific behavior in level 4; I would suggest d).

<small-rant>I think all of this has been proposed and discussed before. The current
state of the spec is not the only outcome I'd be happy with, so I am not objecting
to reopening this topic, but if we're going to reopen to essentially the same set
of options, it'd be nice to hear what the new arguments are, rather than just the
conclusion. </small-rant>

Moaning aside, to clarify:
- a is the current firefox behavior
- b is the current spec behavior
- c is the current webkit/blink behavior
- d is the current IE behavior

The way I see it:
- c and d are clearly alternative attempts at doing something similar. I can see browser vendors wanting either (for platform consistency, inertia), and I have a harder time seeing authors wanting to switch from one to the other (the difference is too small).

- a is a bit like c or d, except it's less good. The spec used to ask for it, and firefox implements it, so I'd be ok with leaving it in as an allowed alternative to c or d, but if firefox is willing to change, we might as well drop it.

- b is different, and makes sure every single character, including spaces, are visible, and as such may be more appropriate when you're dealing with text where white space is important / significant (source code, etc).

Based on that, I'd be ok with:
i - for white-space:pre-wrap & overflow-wrap:normal, allow any of a (unless FF is willing to adjust), c or d (and maybe also b) at (level 3), and recommend d (level 4)

ii - for white-space:pre-wrap & overflow-wrap: break-word, require b (level 3)

iii - introduce overflow-wrap: break-spaces (level 4), also requiring b, but otherwise behaving the same as overflow-wrap: normal.

This is actually closer to my original preference than what we have in the spec
now. See that old mail for more nuances and ascii art diagrams of all the variants:

Anyway, I think making it possible for authors to get the "b" mode is important.
If it's not the one we get by default (current spec), I want a switch (proposal

 - Florian
Received on Monday, 9 November 2015 07:30:47 UTC

This archive was generated by hypermail 2.4.0 : Friday, 25 March 2022 10:08:58 UTC