Re: gtkhtml2 vs. gtkhtml1

[Jeffrey Stedfast <fejj ximian com>]

On Wed, 2002-09-18 at 05:38, Mikael Hallendal wrote:
> tis 2002-09-17 klockan 23.17 skrev Joe Shaw:
> 
> > The bottom line is that it took us two years (in time) to get editing to
> > work as well as it does for Evolution.  And, we had as many as three
> > developers working on it.  So we're talking several man-years of effort
> > here.
> 
> I've tried to get this questions answered a couple of times but never
> really have. *What* was it that took 2 years for three persons? I find
> it very hard to believe that it was only editing, from what I understand
> gtkhtml1 was in a pretty pre-mature and bad state when you guys starting
> hacking on it. 

afaik, the 2 years was definetely spent on more than just editing.

> 
> Also, I imagine that *if* you (or anyone else) where going to add
> editing to GtkHTML2 it would take _a lot_ of less time since lots of the
> efforts from adding it to GtkHTML1 could be reused. And from what I
> understand having the DOM could also help greatly when adding editing.

I've been pondering this too, this definetely sounds like it should be
true (but I don't know the details as I've never hacked on gtkhtml*). Of
course, that doesn't mean adding editing will be a 

"<miguel> 2 day hack!"  :-)

> 
> > I suspect it doesn't much matter whether "we" is the GNOME project or
> > Ximian, but if it's worthwhile to add CSS to gtkhtml1, then there's no
> > real reason not to.  Ximian must have an editable HTML widget for
> > Evolution.  gtkhtml1 ported to gnome 2 makes the most sense for that,
> > since we don't have the time or resources to reimplement it in gtkhtml2.
> 
> I think that Ximian should go with GtkHTML1 since it currently does
> everything they need for Evolution. I think that it would be a waste of
> efforts adding CSS and DOM (not sure if these rumors where true) to
> GtkHTML1 if it's not needed in Evolution. 

Most definetely...

> 
> Preferably someone would step up and say they are interested in making
> GtkHTML2 do editing and everything else we would want from a HTML-widget
> in GNOME. And perhaps Ximian and the Evolution-team could switch to
> GtkHTML2 at that point and we would have only one HTML-widget for the
> GNOME project.
> 
> For a full-blown web browser I think that neither GtkHTML1 or GtkHTML2
> is good enough and we should just do as Galeon has and use Gecko.

I agree. But then again, for my needs - either would be fine ;-)

> 
> > I would too, but even the *original* gtkhtml2 authors aren't hacking on
> > it!  The only contributions lately have been accessibility work by Sun
> > and Wipro and the occassional bug fix.
> 
> Yes, at this point GtkHTML2 is considered a private library for Yelp and
> is not something that people should depend on without knowing this.
> 
> We might want to have *one* official HTML widget that we say, use this,
> we will continue not to break it for you. And make the other one a
> private library for the app/apps using it.

This would be nice.

> 
> At this point I'd go say that GtkHTML2 should be that officially
> supported HTML widget and we make GtkHTML1 a private library of
> Evolution.

Seeing as how gtkhtml2 supports more of the rendering features that
people want as far as I know (and mostly, people are just going to be
using it for rendering anyway), I would have to say that this is
probably a good idea.

> 
> > How do you reconcile this problem?  How do you propose that Evolution
> > users write an email?  Okay, okay, you might argue that no one should
> > send HTML mail, but then consider a WYSIWYG web page editor like
> > Bluefish?  (I don't know if it actually uses gtkhtml1, but it's just an
> > example)
> 
> Btw, Balsa is using GtkHTML2 if I'm not mistaken. How do they solve the
> editing bit? (Perhaps you can only view HTML-mails but not write them?)

probably can't edit html messages.

Jeff

(Note: I do not speak for Ximian, I speak for me)

-- 
Jeffrey Stedfast
Evolution Hacker - Ximian, Inc.
fejj ximian com  - www.ximian.com