Re: component architecture




I'm pretty sure that by now MICO is the ORB of choice, though a new
thought appears to be to create an ORB from scratch, with the central
features that GNOME needs.  But short of that, I think development is
progressing with MICO.

As far as GUI framework is concerned, I think that Owen Taylor is working
on the C bindings for MICO so that GTK+ and the supporting GNOME
widgets (like XmHTML) can be CORBAized. I don't know what stages this is
in. Can someone update on this? 

Once that is done, then I think we can see some real progress towards
building a thorough component infrastructure.  I agree with you that the
creation of many a monolithic applications will cause trouble when the
time comes to componentize the applications for the greatest amount of
integration and reusability.  What it looks like is going to happen is
that we're gonna start building GNOME monolithically and then slowly but
surely rip apart the applications and componentize them.  

I believe that the idea is to show (either as proof-of-concept or
something more) that GNOME is a very feasible and usable idea that's got
the app support built-in and then move from there.  Could one of the
principals please veryify/refute/flame/otherwise on this?

Though i'm not fully comfortable with the idea (and like it matters if i
am!), the drive is probably to get GNOME up to par with KDE as soon as
possible.

I think that once we get the idl compiler and the support in MICO for C
(and some other languages), then we can begin discussion on interfaces and
design/publish a well-concieved set of interfaces for :
	1) interacting with the WM and Panel
	2) interacting with the Network
	3) Interacting with other software (Like audio players, etc)

It's obvious that some applications lend themselves very well to
componentization (like Web Browsers). In a wild dream, we could make
GNOMEE (GNOME Explorer) support component CORBA objects as plugins for all
the stuff that it can do.  In it's entirety, it would just have to be a
restrictive component holder....

Any thoughts?

aadi deshpande	
aadi@bigfoot.com
work://arcus.net
school://njit.edu


On Sat, 21 Feb 1998, Tim Voght wrote:

> Hi,
> 
>  I've been lurking around over the last several days checking out the
> Gnome
> project and trying to determine if it's something I'd want to get
> involved
> with.
>  I'm impressed with the accumulation of stuff and with the online CVS.
> I like the Gtk toolkit.
> The Gnome project seems to have a lot of momentum. 
>  
>  But I need some clarification on the component architecture. I've
> caught up
> on the archives from the the component-list and I see that some are
> concentrating on using MICO as an ORB, and I see a lot of talk
> everywhere
> about the 'component architecture' of Gnome.
> 
>  What's not clear at all is how the ORB is going to be used, and what
> constitutes a component. I would have expected a GUI framework to be a
> central
> element, with the ORB being the central binding element of the
> framework.
> I would have expected applications to be built of relatively
> fine-grained
> components. I have not seen any discussion of a framework. Am I missing
> something? Seems like such a framework would have to be in place before
> you'd
> even start developing components.
> 
> Meanwhile, I see a growing list of apparently monolithic applications...
> 
> Can anybody point out what I'm missing? I haven't had a chance to go
> through
> the CVS. I'm going by what I see on the Web page and in the gnome-list
> and
> component-list.
> 
> FWIW, I have ideas about a very basic GUI framework based loosely on the
> original Model-View-Controller concept introduced by Smalltalk and
> re-used
> by so many other frameworks...
> 
> tim
> 
> 
> -- 
>          To unsubscribe: mail gnome-components-list-request@gnome.org with 
>                        "unsubscribe" as the Subject.
> 
> 




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