On 3/26/07, Quentin Mathé <[EMAIL PROTECTED]> wrote:
Hi Yen-Ju
Le 26 mars 07 à 08:12, Yen-Ju Chen a écrit :
> Some of you might notice that I am working
> on a very simple file manager, called OuterSpace.
> You can find in under etoile/brahcnes/yjchen/OuterSpace.
> It is based on Jesse's design:
> http://jesseross.com/clients/gnustep/ui/concepts/01/ui.png
> But it is obvious that it needs some improvement here and there.
> Help is welcome.
I have taken a look at your commits :-)
I know it is not something people want to see now.
But instead of discussing something not existing,
I think an implementation on hand is more practical for discussion.
Why not use UKDistributedView rather than reimplementing a new icon
view with few features? I know that UKDistributedView is GPL so you
need to put OuterSpace code dual-license and released it as GPL.
Initially it should be fine I think. We could write a new icon view
from scratch later.
My implementation is not going to be
as flexible as UKDistributedView for a while.
But I figure if I start to improve UKDistributedView,
it is all locked in GPL.
Therefore, why not just write one from scratch so that all my effort
will not be locked.
You should also follow code guidelines (specially for indentation) if
we decide to merge it into trunk later.
I do my best. But I think it is not going to be merged anytime soon.
> My plan is to use it as a platform to implement all the low-level
> file operation
If you want to do so, try to do it with a very low coupling from the
rest of the code and in a flexible way, so we can replace this code
by CoreObject when it's ready.
I know we need some kind of solid spec and API for CoreObject.
Yes, we need a CoreObject implementation first.
And seeing that we have so many problems with GWorkspace,
I don't think we will have time to wait for CoreObject.
It will only be used for the next major release, if there is any.
> and hopefully it can turn into a framework for a more advanced file
> manager.
> I attach a screen shot for fun.
> It needs IconKit for icons, and recycler icon is taken from GWorkspace
> (not in svn).
> It is still on early stage.
> None of the file operation is implemented.
> I just finish up the UI.
I'm still evaluating to use Space, it is roughly similar but has
different features (like simple Search, list view), but no Shelf and
not spatiality.
OuterSpace will be probably the most simple file manager
you are going to see. :)
The goal is try to have all the low-level file operation ready.
So by the time Space is ready, it can probably just hook up what we have
and becomes a solid file manager.
I still think Io is better to experiment with file managers ideas.
When I saw you first commit, I was a bit sad because together we
could have finished to fix Io rather than using this time to write a
new file manager from scratch.
I would like to have Io very solid and use it for some applications
(like Spot), that's why I have been taking to time on Io ObjcBridge
and I want it to be fixed.
I mentioned my attitude of Io on SILC once.
I do try to work it out.
But after looking at the change of Io and the effort we need to make it work
with objective-c runtime,
I feel it is too much of work for now.
They often have dramatic changes from time to time.
For example, Vector is gone suddenly, probably replaced by Box.
All the NSRect, NSPoint, NSSize have to be rewritten.
I also mentioned that their ObjC addon does even not work.
File manager is what people reply on.
The stability is the first priority.
The damage may be more than just a crash.
Io may be a good tool to play with idea.
But for the final implementation of a file manager,
I think it is better not to use it until it is stable.
For other applications, such as Calc and Spot,
I am fine with Io, and I think it is not a bad idea.
Finally whatever happens. If we use Space initially, when we decide
to move from prototype stage to real implementation, we could move
from Space to OuterSpace.
If finally I decide to give up on Space, we could work together on
OuterSpace, we should then take time to discuss the design a bit more
to be sure we can achieve our goals.
I believe we are going to have a big discussion about UI.
That is why I am not going to work on the UI more.
What you see in screen shot is probably what it will be in the future.
Again, my purpose is to work out all the implementation first.
I put as much comment as possible in the code.
So in the future, whoever want to write a file manager in any language
can take a look and figure out how to do things,
like preview, drag-and-drop, custom view, desktop entry, trash can, etc.
Which one will be used by Etoile is not my concern
as long as they all use the same underneath frameworks in the end.
UI is just a skin on top of it.
Yen-Ju
Cheers,
Quentin.
_______________________________________________
Etoile-dev mailing list
[email protected]
https://mail.gna.org/listinfo/etoile-dev
_______________________________________________
Etoile-dev mailing list
[email protected]
https://mail.gna.org/listinfo/etoile-dev
