[sword-devel] roadmap for Windows frontend(?)

[Daniel Glassey]

ok, here's my first thoughts to throw into the mixture.

Chris Little wrote:
> I would be interested in getting a feel for where people believe we should
> head with Windows frontend development (after 1.5.6).  I'm really only
> interested in realistic and feasible near-term objectives, especially from
> those people who are likely to actually put work into the project.  The
> reason I ask is because I would like to know where I should devote my
> time.
> 
> As far as I can see, there are four roads we can take with BibleCS at the 
> moment:
> 
> 1. Keep doing the same as we've been doing.
> 
> 2. Move to Gecko rendering with the same BibleCS codebase.
> 
> 3. Develop the desktop-based prototype to match the features of the
> current BibleCS codebase (and presumably add Gecko rendering).
> 
> 4. Dump BibleCS development to concentrate on the wxWindows-based
> frontend.
> -----
> 
> My own hopes/feelings/vision....
> 
> I'm somewhat torn between options 2 and 3.  I feel as if it is really time
> to move past RTF (ruling out option 1)

I definitely agree with that.

> and I feel that wxWindows, while a
> very nice way to support multiple platforms, cannot serve the
> Windows-using community as well as a Windows-specific API would (ruling
> out option 4).  (wxWindows, for example, could not manage the
> desktop-interface of the prototype very easily, to my knowledge.)

If you mean can it do a delphi/bcb style layout like the
biblecs-prototype then afaik the answer is yes (at least wxPython can do
that since boa-constructor (wxPython dev ide in progress) is built like
that), but that would need investigation.

> Personally, I wish we could dump BibleCS and do an MFC front-end in VC++ 
> since that would eliminate a lot of the nonsense we put up with, having to 
> essentially port 3rd party libraries from VC++ to Borland, but I don't 
> think there are enough like-minded developers to support moving 
> development to VC++.

This is where wxWindows has an advantage, it is compiler neutral.

> So I guess my vote, out of the options I listed, would be to do option 2 
> for the next release (get Gecko rendering done) then use that to really 
> get started on option 3: BibleCS 2.0.

I haven't kept track if it but has there been any progress on compiling
gecko on bcb? Or is the aim going to be to compile gecko into a dll
somehow with vc++, then create an import lib for bcb and then use it
that way?
How long do you think it will take to get gecko rendering done?

> Some of the big advantages of finally moving to HTML rendering will be 
> ability to produce interlinear texts that actually stack and to use CSS 
> for user-customization.  With CSS, we could even toggle things like 
> Strong's numbers or footnotes by a change to the CSS instead of running 
> through our filters.

HTML rendering is definitely the way to go.
And, if compiling it is possible and the bloat isn't too much imho gecko 
is obviously the renderer of choice for a GPLd app on windows if full 
html and css are required, if just basic tags are enough then wxHTML is 
sufficient.

fwiw there is a wxMozilla project for embedding mozilla in a wxwindows
app  http://www.theolliviers.com/wxMozilla/ and there has been talk (one 
mail from Robin Dunn on the wxPython list) about OSAF doing a wxGecko 
though it's probably vapourware.

> Thoughts/comments?


hth,
Daniel