[Top][All Lists]

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

RE: [Devel] initial release of FTLayout/GXLayout

From: Turner David
Subject: RE: [Devel] initial release of FTLayout/GXLayout
Date: Wed, 25 Feb 2004 13:15:19 +0100

Hi again,

> Thank you for your concern.
> >   - OTLayout seems to be under LGPL. There is no chance that we're
> >     going to distribute it in the FreeType 2 source package, since
> >     this license is more restrictive than the current dual-license
> >     plan we're using. For example, LGPL code is practically unusable
> >     in most embedded systems.
> Do you mean *I* cannot distribute ftl-20040225.tar.gz by myself?  Or 
> FreeType project cannot distribute a version of FreeType2 including 
> OTLayout? ( Of couser I have to inspect this issue by myself. )
> Anyway I have just remoevd the ftl-* from the web server.
Simply that _we_ will not distribute this code in our official releases,
nor accept it in the CVS. You're free to distribute any code you like
(as long as you agree to the terms of the code you didn't authored).

Basically, this means that:

  - you're free to distribute a mix of GPL/LGPL/FTL code in a
    single source archive (i.e. there is no need to remove the
    file from your webserver)

  - as far as I know, the FTL is incompatible with the GPL (hence the
    dual-license plan). I don't know exactly if it is compatible with
    the LGPL as well. If this is *not* the case, the only way to legally
    distribute *binaries* of the resulting library as a single
    file is under the GPL !

But there is hope, if the LGPL code can be easily replaced by something
different, we could integrate it in our CVS.

And we can still integrate the "GXLayout/FTLayout" stuff once we've
checked it though. It seems the problem if very local to the "OTLayout"

> >   - Wouldn't be better if these components were provided separately
> >     from FreeType ? I have several reasons to believe it'd be
> >     beneficial:
> >
> >        - smaller size of the font engine for those who don't want
> >          to use the extra features
> We can make it small by adding a build option.
> Current GXLayout implementation loads all GX tables into structures.
> We know this is wrong. However, we need to do so to learn about GX
> tables. In the future we will change the code to load table 
> header only.
> We will also add table verification code at loading time.
> We have implemented GXLayout and OTLayout as font drivers. We know
> it is wrong design now. We should extend TrueType and CFF font drivers
> to support GX and OT.(You may agree with us about this issue).

Well, these are implementation details, and it's a bit hard to discuss
about them without deep knowledge of the source code. Rest assured that
we'll find a way to solve these problems together.

> In such case separating the source FT2 and FTLayout will be not good.
> >        - different release points, so that a bug-fix in the layout
> >          engine doesn't generate a new release for the font engine
> I have had no answer yet. This is just initial release.
Well, I'm concerned about support issues. I would appreciate if we wouldn't
have to answer every question regarding this layout engine if we can't answer
them very well. Are you ready to become a "full" member of the FreeType team ?
You'd be welcome !! :-)

> >        - independent testing / debugging.
> >
> >        - the ability to use a different font engine !
> I don't think so. Our primary intent is writing a text layout
> engine on FreeType2. I don't have much interest in other font 
> engine. Is there any good font engine other than FT2? Instead 
> we want to write a text layout engine using FT2's facilities: 
> memory management, stream, service and resource fork accessors:-P
OK, I'll try to answer specifically to this point later. 

> I think your approach about old ftlayout/otlayout is wrong.
Could you care to elaborate ?

> Many good codes for typography hacking are in FreeType2. We should use
> it, shouldn't we?
> See also:
> >        - independent license (LGPL if you want)
> No, I don't want. FTLayout/GXLayout is written from scratch. 
> So we can apply the FT2's dual-license plan. OTLayout is the main
> target of this post. If needed, Of course, I will remove from
> ftl tar.gz file. I have no time to estimate the time to rewrite
> LGPL part of OTLayout.
OK, I'll give you an estimate of what we can do after reading the
source code. I suppose that there are a lot of similarities with
the "otlayout" module on the CVS anyway.

> > For the name, I'm proposing these alternatives names that don't use
> > the FreeType moniker:
> > 
> >   - FontScript  (because the package is really about 
> international scripts)
> > 
> >   - TypeScript  (same idea here)
> > 
> >   - ScriptText
> > 
> >   - Semtex  (it doesn't need to carry exact meaning, does it ;-)
> > 
> >   - RedHat Layout (if you really insist)
> If you, ft developers permit to use, we'd like to use the 
> name FTLayout. Because our text layout engine is not ICU; our code highly 
> depends on
> FT2. Even if it is not permitted to use FTLayout, we would like to use

I'd like to give the "FreeType Layout" name to a library that provides
a rather high-level text layout facility to application developers. Something
that looks more like Apple's ATSUI than Pango. I need to check that what
you provide corresponds to this; I'm also a bit anal on API abstraction
sometimes :-)

> - FreeLayout or
> - FreeScript
> We will use "Free" as prefix.
OK, no problem with that.

Give me one week to review the code completely. Unfortunately, I doubt I'll
be able to do something useful with it before that.

By the way, would you allow me to integrate the non-OTLayout code into
the CVS (possibly with minor corrections) as found in your latest archive ?


- David Turner
- The FreeType Project  (

This message and any attachments (the "message") is intended solely for the
addressees and is confidential. If you receive this message in error, please
delete it and immediately notify the sender.
Any use not in accordance with its purpose, any dissemination or disclosure,
either whole or partial, is prohibited except formal approval.
The E-Mail transmission can not guarantee the integrity of this message.
NDS TECHNOLOGIES FRANCE will not therefore be liable for the message if 

reply via email to

[Prev in Thread] Current Thread [Next in Thread]