Pango and SILGraphite was Re: Possible Pango 1.4 ideas

["Daniel Glassey" <danglassey ntlworld com>]

Hi,
After hearing about the discussion of ideas for pango 1.4 a 
discussion has started on the silgraphite list about the possibility 
of integrating graphite into pango.

SILGraphite (graphite.sil.org) is a library for providing "rendering 
capabilities for complex non-Roman writing systems"/"extensible 
complex script behavior"

It is written in C++ but C wrappers for use in C-based projects are 
being considered - so it would have to be some kind of plugin - it 
isn't for a single writing system and is just for all graphite-
compiled fonts so I'm not sure if it will fit into the current 
engine/plugin system.

It is currently being ported to linux 
(www.sf.net/projects/silgraphite) and being integrated into windows 
Mozilla and hopefully others later (sila.mozdev.org)

Would this be feasible/useful for 1.4? We need to work on a detailed 
API proposal and patches , but is it conceptually the kind of thing 
you would be interested in? also, see summaries below.

Regards,
Daniel

P.S. where has www.pango.org gone??

_______________________________________________
Silgraphite-devel mailing list
Silgraphite-devel lists sourceforge net
http://lists.sourceforge.net/lists/listinfo/silgraphite-devel

On 22 Jan 2003 at 10:59, Sharon_Correll sil org sent forth the message:

[other bits]
> Still, I think Graphite is a much more elegant solution to the problem of
> extensible complex script behavior. And certainly much more practical for
> organizations like SIL,especially if we are going to be using Graphite on
> Windows anyway. (I'd like to think that developing a critical mass of those
> who can tweak/extend GDL would be easier than developing a comparable body
> of people who can tweak/extend Pango! I hope I'm right!) And I've talked to
> enough people at conferences to get the impression that there are those in
> the Linux community who feel that Graphite would be filling a hole.


On 22 Jan 2003 at 15:10, Peter_Constable sil org sent forth the message:

> On 01/22/2003 10:35:19 AM "Daniel Glassey" wrote:
> 
> >Basically, I'm a bit unclear as to what the hole it would fill is, or
> >at least what resources like fonts and things there will be that will
> >require that hole to be filled. Without that a Graphite-Pango
> >connection is imho unlikely.
> 
> My knowledge of *nix is quite limited, but I've had the impression that
> there is a hole that is worth filling. (As Sharon said, there have
> certainly been people at conferences over the past few years saying this.)
> Here's my understanding:
> 
> If we're talking about support for internationalised text, we're talking
> about some kind of graphics subsystem used for application development
> (whether it is part of a comprehensive API set that provides many kinds of
> services, or something providing only graphics-related services). My
> understanding is that in the *nix world there are various graphics systems
> that might be used. To support international text, any of these has to
> incorporate support for complex script behaviours. My understanding is that
> Pango was intended to be a subsystem that might be used by a variety of
> different graphics systems to deal with this specific task. (It's not clear
> to me how STSF and ICU fit into this picture.)
> 
> To the extent, then, that one or more graphics systems rely on Pango to
> deal with complex-script layout, any improvements to Pango in this regard
> will potentially benefit applications that use those graphics systems.
> (It's not clear to me whether an application can utilise Pango while also
> working on a graphics subsystem that itself does not utilise Pango; i.e.
> make calls to Pango to get from strings of characters to a sequence of
> segments each consisting of a sequence of glyphs with positioning
> information.)
> 
> Now, it's my understanding that Pango has been architected so that
> rendering support for writing systems is implemented one writing system per
> module, meaning that supporting any new writing system requires writing
> code specifically for that writing system. That's the kind of extensibility
> bottleneck that Graphite was specifically designed to overcome. There ought
> to be a way to incorporate Graphite into Pango such that any writing system
> can be supported by Pango by simply installing an appropriate
> Graphite-enabled font.
> 
> If my understanding is correct, then, the potential for Graphite to be
> filling a hole if incorporated into Pango is equal to a combination of the
> extent to which Pango does not support all writing systems together with
> the extent that Pango is utilised by different graphics systems (or
> directly by applications, if that is possible) to provide complex-script
> layout capabilities.
> 
> 
> Hope I haven't stated the obvious.
> 
> - Peter
> 
> 
> ---------------------------------------------------------------------------
> Peter Constable
> 
> Non-Roman Script Initiative, SIL International
> 7500 W. Camp Wisdom Rd., Dallas, TX 75236, USA
> Tel: +1 972 708 7485