I18N-ISSUE-167

Example of Bidirectional Text http://dev.w3.org/csswg/css3-writing-modes/#bidi-example I've been meaning to say for some years now that this is a very bad example. It should use dedicated bidi specific markup (see http://www.w3.org/International/questions/qa-bidi-css-markup#markup "You should therefore use dedicated bidi markup whenever it is available. Do not simply attach CSS styling to a general element to achieve the effect."). Here is an proposal for an alternative version of parts of the example. <HEBREW dir="rtl"> <PAR>HEBREW1 HEBREW2 english3 HEBREW4 HEBREW5</PAR> <PAR>HEBREW6 <EMPH>HEBREW7</EMPH> HEBREW8</PAR> </HEBREW> <ENGLISH dir="ltr"> <PAR>english9 english10 english11 HEBREW12 HEBREW13</PAR> <PAR>english14 english15 english16</PAR> <PAR>english17 <HE-QUO dir="rtl">HEBREW18 english19 HEBREW20</HE-QUO></PAR> </ENGLISH> /* Rules for bidi */ *[dir=rtl] {direction: rtl; unicode-bidi: embed;} *[dir=ltr] {direction: ltr; unicode-bidi: embed;} These generic style rules should be what goes in the separate style sheet. Also, we recommend not using bidi markup unless you need to change the base direction, so if this document had an overall base direction of ltr (either by default, or via <ROOT dir="ltr">), you wouldn't need to have the dir="ltr" after ENGLISH. It may be worth adding a note to that effect. It's also confusing that the markup is in uppercase, since the uppercase is used to indicate Hebrew characters. Unless you are trying to make a point that the markup uses hebrew element names (which I don't think is necessary here), they should probably be in lowercase. -- Richard Ishida Internationalization Activity Lead W3C (World Wide Web Consortium) http://www.w3.org/International/ http://rishida.net/

On 10/17/2012 10:01 AM, Richard Ishida wrote: > Example of Bidirectional Text > http://dev.w3.org/csswg/css3-writing-modes/#bidi-example > > I've been meaning to say for some years now that this is a very bad example. It should use dedicated bidi specific markup (see > http://www.w3.org/International/questions/qa-bidi-css-markup#markup "You should therefore use dedicated bidi markup whenever > it is available. Do not simply attach CSS styling to a general element to achieve the effect."). Here is an proposal for an > alternative version of parts of the example. I switched the example to use DocBook. Take a look and let me know if it's good? > /* Rules for bidi */ > *[dir=rtl] {direction: rtl; unicode-bidi: embed;} > *[dir=ltr] {direction: ltr; unicode-bidi: embed;} > > These generic style rules should be what goes in the separate style sheet. Not sure I handled this part as you want; I think you want me to adjust the prose somehow? > Also, we recommend not using bidi markup unless you need to change the base direction, so if this document had an overall base > direction of ltr (either by default, or via <ROOT dir="ltr">), you wouldn't need to have the dir="ltr" after ENGLISH. It may > be worth adding a note to that effect. I think it's not that important to point out here. > It's also confusing that the markup is in uppercase, since the uppercase is used to indicate Hebrew characters. Unless you are > trying to make a point that the markup uses hebrew element names (which I don't think is necessary here), they should probably > be in lowercase. Good point. Fixed. ~fantasai

Hi Elika, On 26.11.2012 22:35, fantasai wrote: >> I've been meaning to say for some years now that this is a very bad >> example. It should use dedicated bidi specific markup (see >> http://www.w3.org/International/questions/qa-bidi-css-markup#markup >> "You should therefore use dedicated bidi markup whenever >> it is available. Do not simply attach CSS styling to a general element >> to achieve the effect."). Here is an proposal for an >> alternative version of parts of the example. > > I switched the example to use DocBook. Take a look and let me know if > it's good? I think that it would be better to use XML and not SGML for DocBook example. Almost noone is using SGML version of DocBook now, except for few legacy projects. So changing SGML->XML and putting attribute values into quotes should make the trick. Also in DocBook each section must have <title> element as a first child. Otherwise I don't see any other DocBook issues in this example. Jirka -- ------------------------------------------------------------------ Jirka Kosek e-mail: jirka@kosek.cz http://xmlguru.cz ------------------------------------------------------------------ Professional XML consulting and training services DocBook customization, custom XSLT/XSL-FO document processing ------------------------------------------------------------------ OASIS DocBook TC member, W3C Invited Expert, ISO JTC1/SC34 rep. ------------------------------------------------------------------ Bringing you XML Prague conference http://xmlprague.cz ------------------------------------------------------------------

Just looked at the latest CR draft http://www.w3.org/TR/2014/CR-css-writing-modes-3-20140320/ I think that the CSS in example 2: /* Rules for bidi */ [dir=ltr] {direction: rtl;} [dir=rtl] {direction: ltr;} quote {unicode-bidi: isolate;} should probably now be: /* Rules for bidi */ [dir=ltr] {direction: rtl; unicode-bidi: isolate; } [dir=rtl] {direction: ltr; unicode-bidi: isolate; } This concords with the advice we give authors in http://www.w3.org/International/articles/inline-bidi-markup/ and with the shift to dir being always isolating, and it also makes the extra line for quote unnecessary. RI On 26/11/2012 21:35, fantasai wrote: > On 10/17/2012 10:01 AM, Richard Ishida wrote: >> Example of Bidirectional Text >> http://dev.w3.org/csswg/css3-writing-modes/#bidi-example >> >> I've been meaning to say for some years now that this is a very bad >> example. It should use dedicated bidi specific markup (see >> http://www.w3.org/International/questions/qa-bidi-css-markup#markup >> "You should therefore use dedicated bidi markup whenever >> it is available. Do not simply attach CSS styling to a general element >> to achieve the effect."). Here is an proposal for an >> alternative version of parts of the example. > > I switched the example to use DocBook. Take a look and let me know if > it's good? > >> /* Rules for bidi */ >> *[dir=rtl] {direction: rtl; unicode-bidi: embed;} >> *[dir=ltr] {direction: ltr; unicode-bidi: embed;} >> >> These generic style rules should be what goes in the separate style >> sheet. > > Not sure I handled this part as you want; I think you want me to adjust > the prose somehow? > >> Also, we recommend not using bidi markup unless you need to change the >> base direction, so if this document had an overall base >> direction of ltr (either by default, or via <ROOT dir="ltr">), you >> wouldn't need to have the dir="ltr" after ENGLISH. It may >> be worth adding a note to that effect. > > I think it's not that important to point out here. > >> It's also confusing that the markup is in uppercase, since the >> uppercase is used to indicate Hebrew characters. Unless you are >> trying to make a point that the markup uses hebrew element names >> (which I don't think is necessary here), they should probably >> be in lowercase. > > Good point. Fixed. > > ~fantasai >

On 26/11/2012 21:35, fantasai wrote: > On 10/17/2012 10:01 AM, Richard Ishida wrote: >> Example of Bidirectional Text >> http://dev.w3.org/csswg/css3-writing-modes/#bidi-example >> >> I've been meaning to say for some years now that this is a very bad >> example. It should use dedicated bidi specific markup (see >> http://www.w3.org/International/questions/qa-bidi-css-markup#markup >> "You should therefore use dedicated bidi markup whenever >> it is available. Do not simply attach CSS styling to a general element >> to achieve the effect."). Here is an proposal for an >> alternative version of parts of the example. > > I switched the example to use DocBook. Take a look and let me know if > it's good? Much better. Thanks. >> /* Rules for bidi */ >> *[dir=rtl] {direction: rtl; unicode-bidi: embed;} >> *[dir=ltr] {direction: ltr; unicode-bidi: embed;} >> >> These generic style rules should be what goes in the separate style >> sheet. > > Not sure I handled thispart as you want; I think you want me to adjust > the prose somehow? I think it's ok, thanks. However the indent after the first line of those paragraphs looks a bit odd to me - was it intentional? I think i'd also make the explanatory text proportionally-spaced, if it was up to me (like the first para of the example). > >> Also, we recommend not using bidi markup unless you need to change the >> base direction, so if this document had an overall base >> direction of ltr (either by default, or via <ROOT dir="ltr">), you >> wouldn't need to have the dir="ltr" after ENGLISH. It may >> be worth adding a note to that effect. > > I think it's not that important to point out here. Ok. > >> It's also confusing that the markup is in uppercase, since the >> uppercase is used to indicate Hebrew characters. Unless you are >> trying to make a point that the markup uses hebrew element names >> (which I don't think is necessary here), they should probably >> be in lowercase. > > Good point. Fixed. Thanks. One further editorial change that may help is to move the following text immediately after the CSS code: [[ The formatting of this text might look like this if the line length is long: ... Note that the <quote> embedding causes HEBREW18 to be to the right of english19. ]] That would make it easier to understand when you say things like "Thus, the first two <para>s are read starting at the top right, the final three are read starting at the top left", because at the moment that appears to refer to the source code, which is in logical order, and so is rather confusing. If you don't want to move it, maybe add something like "in the displayed text" to explain where the reader should be looking. One final suggestion: do we need to have hyphenation in the lower example of displayed text? Also the right-aligned lines are not properly right-aligned. RI PS: I still have another comment in a separate email about adding unicode-bidi: isolate to the CSS. And Jirka made a couple of suggestions about the DocBook syntax.

On 03/24/2014 09:11 AM, Richard Ishida wrote: > Just looked at the latest CR draft http://www.w3.org/TR/2014/CR-css-writing-modes-3-20140320/ > > I think that the CSS in example 2: > > /* Rules for bidi */ > [dir=ltr] {direction: rtl;} > [dir=rtl] {direction: ltr;} > quote {unicode-bidi: isolate;} > > > should probably now be: > > /* Rules for bidi */ > [dir=ltr] {direction: rtl; unicode-bidi: isolate; } > [dir=rtl] {direction: ltr; unicode-bidi: isolate; } > > > This concords with the advice we give authors in http://www.w3.org/International/articles/inline-bidi-markup/ and with the > shift to dir being always isolating, and it also makes the extra line for quote unnecessary. Fixed. ~fantasai

On 06/17/2014 07:39 AM, Richard Ishida wrote: > > I think it's ok, thanks. However the indent after the first line of > those paragraphs looks a bit odd to me - was it intentional? I think > i'd also make the explanatory text proportionally-spaced, if it was > up to me (like the first para of the example). Hmm. Looks like the spec postprocessor is choking on something. Should be worked around now (and I'll need to file a bug). > One further editorial change that may help is to move the following > text immediately after the CSS code: > > [[ > The formatting of this text might look like this if the line length is long: > ... > Note that the <quote> embedding causes HEBREW18 to be to the right of english19. > ]] > > That would make it easier to understand when you say things like > "Thus, the first two <para>s are read starting at the top right, > the final three are read starting at the top left", because at > the moment that appears to refer to the source code, which is in > logical order, and so is rather confusing. Okay, done. Also did a bit of related cleanup. > Also the right-aligned lines are not properly right-aligned. Right. Appears to also be a bug in the postprocessor. Added a workaround to this one, too, so should be fixed now. > One final suggestion: do we need to have hyphenation in the > lower example of displayed text? Um. No. But it needs to go somewhere, so I'm going to leave this issue open until that concern lands elsewhere. Thanks for the detailed feedback and good suggestions! ~fantasai