[mpeg-OTspec] rclt feature -- minor concern

[Behdad Esfahbod]

On 08/08/2012 07:20 PM, Mansour, Kamal wrote:
>  
> Sairus,
> 
> To put it in perspective, I would say that the 'rclt' feature is a welcome
> addition for the more complex styles of Arabic writing, but it would not be
> required for many simple styles. If it's present in a font, it should not be
> normally possible to turn it off. By the way, 'rclt' will also be useful to
> many other scripts which might require  more than one pass to refine the final
> rendered shapes. 

So, why can't everyone simply turn 'calt' on by default?  That's what Pango /
HarfBuzz / etc have been doing.  Not that I oppose 'rclt' (I already activated
it in HarfBuzz).  I just think we already have more than enough features.
Specially, features that are not supposed to be user-controllable can all be
gathered under one feature, there is no need to add more tags.

As for your "one pass" comment, that's invalid.  A single feature can contain
as many lookups, ie. passes, as needed to do its job.  Or did I misunderstand you?

My 0.02CAD,
behdad


> Kamal
> 
> From: Sairus Patel <sppatel at adobe.com <mailto:sppatel at adobe.com>>
> Date: Wed, 8 August 2012 14:38
> To: "mpeg-OTspec at yahoogroups.com <mailto:mpeg-OTspec at yahoogroups.com>"
> <mpeg-OTspec at yahoogroups.com <mailto:mpeg-OTspec at yahoogroups.com>>
> Subject: [mpeg-OTspec] rclt feature -- minor concern
> 
>  
> 
> Folks,
> 
>  
> 
> 1. I received feedback that in the following sentence from the rclt specification:
> 
>  
> 
>> This feature is similar to 'calt', but with the difference that it should not
> be possible to turn off 'rclt' substitutions: they are considered essential to
> correct layout of the font.
> 
>  
> 
> “it should not be possible to turn off 'rclt'” is a bit too strong. There are
> always workflows in which the author may need to turn off even “required”
> features e.g. when preparing a specimen book chart, or other special-case
> scenarios.
> 
>  
> 
> I looked at the specification for ‘rlig’ and don’t see any such strong
> wording. So I propose the following to replace the above quoted sentence:
> 
>  
> 
>> This feature is similar to 'calt', but with the difference that 'rclt'
> substitutions are considered essential to correct layout of the font.
> 
>  
> 
> The recommendation that this feature “not be turned off, to avoid breaking
> obligatory shaping” is already in the UI suggestion of the rclt specification
> proposal.
> 
>  
> 
> 2. Speaking of ‘rlig’, are the Arabic (and other) layout specs going to be
> updated to include ‘rclt’ as a required feature? rclt sounds like it has the
> same level of “requiredness” as rlig.
> 
>  
> 
> Best,
> 
> Sairus
> 
>  
> 
>  
> 
> *From:*mpeg-OTspec at yahoogroups.com <mailto:mpeg-OTspec at yahoogroups.com>
> [mailto:mpeg-OTspec at yahoogroups.com] *On Behalf Of *Levantovsky, Vladimir
> *Sent:* Monday, August 06, 2012 1:55 PM
> *To:* Levantovsky, Vladimir; Thomas Phinney; Michelle Perham; OTspec
> *Subject:* RE: [mpeg-OTspec] RE: New work on 3rd edition of the OFF (AHG kick-off)
> 
>  
> 
>  
> 
> Folks,
> 
>  
> 
> I just uploaded an updated version with the description of ‘rclt’ feature
> added:
> http://groups.yahoo.com/group/mpeg-OTspec/files/20120806-w12847_14496-22_WD1_3rd_ed.zip
> 
>  
> 
> Please note that the editing period for the 3^rd edition working draft text is
> open until the end of August, which means that if there are changes you want
> to propose and those changes have a consensus approval of the group – we can
> add them in the draft. After the editing period is over the changes would need
> to be submitted as input contributions for discussion at the next WG11 meeting
> so I strongly recommend to consider as much things done as possible until the
> end of the month.
> 
>  
> 
> Thank you,
> 
> Vlad
> 
>  
> 
>  
> 
>