So, what should we do about Persist::getIId?



Hi,

I am not really happy that the initial discussion about getIId/getMoniker
between Michael and me occured on iRC instead of a mailing list, so I will
first try to summarize said discussion.

Michael's basic idea is to create a moniker that creates `empty'
(not-yet-loaded-by-any-previously-saved-persist-data) components for a
given mime type, but is smarter than doing just that:

<michael> I'd like to migrate to:
<michael> "handler:application/x-png from gnumeric"
<michael> and have a nicer customization / mime-type <-> component mapping
app that could customize these.
<michael> ie.
<michael> you might want gimp inside gnumeric, but eog inside nautilus.
<michael> or "handler:application/x-png for compound_doc" eg.

His goal is for compound documents to be able to not depend on any given
specific component but at the same time contain more information on what
component to use than the mime type of the embedded data.


My view on this is that this is all great but it doesn't belong in
Persist. This should be done on the container side and not the containee
-- I'm not sure the containee even knows how it was requested (which is
basically the information needed to implement Michael's idea of
Persist::getMoniker). Persist::getIId, in my opinion, should be a
non-compilcated simple method whose sole purpose is to make it possible to
_completely_ save a component (which should be Persist's goal anyway).

Now for a meta-argument: Michael's proposal is (apologies if this is not
the case and there are working implementations already) a non-exising
solution to problems we can't really articulate yet. 
I, for one, love ideas and design discussions that look into the long
term, but it is the 18th of July, I haven't written a single line of
Bonobo-Clipboard code yet, and while it will be (of course) developed
outside Bonobo proper, we need a working change to Persist by the date of
the API freeze.


Please comment,
	Cactus

-- 
   .--= ULLA! =---------------------.   `We are not here to give users what
   \     http://cactus.rulez.org     \   they want'  -- RMS, at GUADEC 2001
    `---= cactus cactus rulez org =---'
A gargarizálás kitűnő módja annak, hogy megállapítsuk nem szivárog-e a nyakunk





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