* Dave Harris <brangdon@cix.compulink.co.uk> [2004-12-23 17:46]:
In-Reply-To: <cqcjii$hs0$1@sea.gmane.org> dave@boost-consulting.com (David Abrahams) wrote (abridged):
That's okay for simple things, but for any serious printing work you need to know you're talking to a printer. For example, you may need to embed postscript in your output stream.
Not all printers support postscript, so you'd need to know not just that you were printing, but what kind of printer it was. Although there probably needs to be some kind of abstraction-busting pass-through mechanism, I see it as the exception rather than the rule.
Another issue is transparency. My employer's drawing framework supports semi-transparency in the form of an alpha channel. This is used by us for things like anti-aliasing, and by users for subtle colour blends. Some printers can support transparency directly and some cannot. We found we needed a relatively elaborate architecture to hide that detail from application drawing code.
I'm a bit concerned about the scope of this boost project. Our drawing framework is pretty huge.
This is an important discussion to have sooner than later. 1) Is a GUI project a bad idea? 2) Is Boost the right place for the a GUI project? * No * no.a) Where does this project belong? * Yes * yes.a) How is a GUI project different from existing Boost projects? yes.b) Which Boost projects have faced challenges similiar to Boost GUI? yes.c) What resources need to be created to support Boost GUI? yes.d) What procedures need to be created to support Boost GUI? I believe that Boost is the right place for a GUI project. I am ready for a new, hybrid approach to GUI development. Q - 1) Yes. I believe the time is ripe for a small, light-weight, XML + CSS renderer to attack the new surge of RSS content on the web. Existing CSS renders are either to big, poorly documented, or are licensed in ways that displease me. I consider how I'd approach the construction of such a component, and I immediately turn to modern C++. I imagine benefits like performance, economy of experssion through generics, conceptual compatibility with underlying OS code. Q - 2) Yes. If the project is to develop a GUI library using modern C++, then Boost is the right place. Boost has the community. Boost has the knowledge. Developing a GUI library is a complex undertaking. Startging out with a new community slows things down considerably. Building a community takes time. Establishing mailing lists takes time. Setting up a web site takes time. Choosing a name takes time. Upon success of the project, it is likely to need a separate umberella, but until then, building a community to support a project so complex, to me, is far more daunting than the software itself. -- Alan Gutierrez - alan@engrm.com