Re: [fltk.general] New HTML Widget
I don't know amaya render but between tkhtml3 and dillo the tkhtml3 has a much better render than dillo. ___ fltk mailing list fltk@easysw.com http://lists.easysw.com/mailman/listinfo/fltk
Re: [fltk.general] New HTML Widget
Just curious - has anyone been doing any work recently on an improved FLTK html widget? A quick trial of the existing help viewer widget indicates it's not likely to support enough HTML for my proposed uses, which raises the same general question already posed in this thread. Tkhtml3 is indeed a self-contained HTML rendering widget that has no external dependencies to speak of (beyond Tcl/Tk itself, obviously.) Converting it to FLTK is almost certainly a lot of work - one first has to understand what Tkhtml is doing (it's about 40,000 lines of code, and more than 2000 of them call things prefixed with Tcl_ or Tk_). It does have the considerable advantage of being license compatible with FLTK. Amaya - the W3C browser/editor is also licensed in a compatible fashion (I think - not sure where the actual rendering parts live), but it isn't immediately clear how much its dependencies are hooked into its core HTML renderer. Amaya has changed GUI toolkits at least once in its history that I can remember (it's currently using wxWidgets) so it's presumably possible to retarget it, but I have no real idea how much work it would be. My guess would be it's at least as much work as tkhtml would be. Webkit is certainly full featured, but it's sheer size and (probable) list of requirements on various platforms makes it a rather daunting prospect. It also uses LGPLv2 with (AFAIK) no static linking exception, which may also pose a problem. Dillo would in some ways be nearly ideal, since they already use FLTK, but they are GPL licensed and thus not currently a candidate. Has anyone ever approached the Dillo devs and asked if they would be willing to support relicensing the parts of their codebase necessary to create a CSS supporting Fl_Html rendering widget so it could become part of FLTK proper? They may or may not be interested in that possibility (it usually depends on the project and the goals of its devs) but if someone who could represent FLTK were willing to ask that would be a great first step. Cheers, CY ___ fltk mailing list fltk@easysw.com http://lists.easysw.com/mailman/listinfo/fltk
Re: [fltk.general] New HTML Widget
On 11/17/11 13:07, Domingo Alvarez Duarte wrote: I was looking again at http://tkhtml.tcl.tk/index.html and found that they did a great advance with their HTML engine that is basically pure C code with a bit of tcl help, this html engine could be a great start point to make a very good html widget for FLTK. Probably a good way to approach this is for someone to make a standalone FLTK widget as a proof of concept, and let that project mature, then we can pull it into FLTK. This approach seemed to work for me with Fl_Input_Choice, Fl_Table, Fl_Native_File_Chooser, Fl_Tree, etc. ___ fltk mailing list fltk@easysw.com http://lists.easysw.com/mailman/listinfo/fltk
Re: [fltk.general] New HTML Widget
On 17 Nov 2011, at 21:07, Domingo Alvarez Duarte wrote: I was looking again at http://tkhtml.tcl.tk/index.html and found that they did a great advance with their HTML engine that is basically pure C code with a bit of tcl help, this html engine could be a great start point to make a very good html widget for FLTK. How usable would the Dillo code be? That might be a good basis, and at least we know it works nicely with fltk! ___ fltk mailing list fltk@easysw.com http://lists.easysw.com/mailman/listinfo/fltk
Re: [fltk.general] New HTML Widget
On 11/17/11 13:56, Ian MacArthur wrote: On 17 Nov 2011, at 21:07, Domingo Alvarez Duarte wrote: I was looking again at http://tkhtml.tcl.tk/index.html and found that they did a great advance with their HTML engine that is basically pure C code with a bit of tcl help, this html engine could be a great start point to make a very good html widget for FLTK. How usable would the Dillo code be? That might be a good basis, and at least we know it works nicely with fltk! I imagine a big issue to watch for is external lib dependencies. I could see where the more fully implemented an HTML browser might be, the more dependencies it will have on a wide variety of external libs. I think any such widget we include in FLTK as a built-in should either be entirely self contained and unbloated (ie. would not double FLTK's source code size), or an option (like Cairo). Without looking at any code, it seems 'webkit' is a good choice on the 'full implementation' end of the spectrum. It has neat extensions that make it usable in standalone apps, and seems to be the way things are going for high end apps. (Not exactly our mission goal, but there it is) I'm guessing webkit also has a very large spanning tree of dependencies, the likes of which would drive most cross-platform devs to self-defenestrate. Offhand: http://trac.webkit.org/wiki/BuildingGtk I think GTK+ and QT have both embraced webkit, as well as Apple (Safari, iPhones, iPads, etc) ___ fltk mailing list fltk@easysw.com http://lists.easysw.com/mailman/listinfo/fltk