Re: Bonobo Clipboard, try #1

[ERDI Gergo <cactus cactus rulez org>]

On Sat, 26 May 2001 textshell t-online de wrote:

> I just want to spread my .02p on clipboards.

You raise some excellent points.

> 	  => we would like to arrange things in a way that those
> 	     apps can manage the data as long as they feel
> 	     comfortable and then (maybe configurable) hand the
> 	     data over to the clipboard server

Yes, this is why the real clipboard is only passing monikers around. If
your app has large datasets and some kind of rollback facility, you are
encouraged to implement your own clipboard state store. But if you are
doing something simple, with small datasets, you don't have to implement
it because you can use the clipboard store in libbonobo, and it also takes
care of the X clipboard for you. The point is flexibility.
(BTW, should the X clipboard also be the responsibility of the app, or
the bonobo clipboard? I would say the former, because if the bonobo
clipboard has to resolve the moniker to PersistStream just to save it to
text/plain, we're pretty much back to square one)

> 	- There should be some reliable and user known way to
> 	  get rid of all the data that is in the clipboard
> 	  (server or still being maintained by the app). I think
> 	  this important to clean up the memory

I'd imagine the clipboard to unref both the pasting and the linking
moniker when the clipboard content changes -- i.e. you would just set
either of them to CORBA_OBJECT_NIL and the clipboard is empty.

> 	- maybe we would like to maintain a clipboard history,
> 	  no need to have it now, but maybe our approach
> 	  should make such things possible.
> 	  (Windows even had/has the option of saveing the clipboard
> 	   to a file)

I believe this can be implemented in this framework solely by the
clipboard implementation.

> 	- Multiple data formats

Monikers give us this for free.

> 	- preview of clipboard contents

You could create an application that watches for clipboard changes (via a
Listener interface connected to the desktop context), then displays it
somehow (maybe by first tyring to resolve it to Control, then to
Embeddable, then if this fails too, to PersistStream and saving it to a
temporary stream in a format it can display (pictures or text are the
obvious possibilities))

-- 
   .--= ULLA! =---------------------.   `We are not here to give users what
   \     http://cactus.rulez.org     \   they want'  -- RMS, at GUADEC 2001
    `---= cactus cactus rulez org =---'
Life is complex;  It consists of real and imaginary parts!