Re: Remove search bar code cruft?

[Curtis Hovey <sinzui cox net>]

On Fri, 2004-03-12 at 08:44, Christian Neumair wrote:
> On Fri, 2004-03-12 at 14:17 +0100, Alexander Larsson wrote:
> 
> > On Fri, 2004-03-12 at 14:01, Christian Neumair wrote:
> > > nautilus_is_search_uri is TRUE (as far as I know we have no search vfs
> > > location at the moment so this is never TRUE), we could simply remove
> > > the whole search capabilities/medusa integration, drastically reducing
> > > compile time. This includes removing the search list view. It would be
> > > enough to bring up an additional gnome-search-tool window that is
> > > triggered through a "Search file or folder..." menu entry. Medusa
> > > integration could be done in gnome-search-tool, since displaying files
> > > that are located in various folders inside a spatial isn't really good,
> > > right?
> > 
> > Compile times would not be affected much at all, certainly nowhere near
> > "drastically". Almost all the medusa-related code is ifdefed out, and is
> > not built.

If you do have Medusa installed, Nautilus will try to compile it and the
build will fail.  The search code still looks for a search service, uses
GTK 1.4 widgets, and is unaware of the icon/theme changes.  

My Nautilus 2.4 patch is terribly broken by spatial and extensions in
2.5; I cannot build Nautilus with Medusa right now.
 
> > We're not yet sure how medusa integration will work. But this requires
> > some serious thinking, and I doubt launching gnome-search-tool is what
> > we want.
> 
> Why? In my opinion integrating this into the file manager doesn't have
> any advantage over having a tool that is well-designed. I know,
> gnome-search-tool currently isn't, but that's not the point.

If you try msearch-gui that comes with Medusa, you'll find yourself
wishing that the files behaved like those in nautilus.  All the items in
the context menu are needed to interact with the rest of the desktop,
with Nautilus.

> > The medusa code is sort of old and bitrotted, and at some point it might
> > be better to remove it and start from scratch. However, I'm not sure
> > we're at that point yet, and I'm not sure there is any way to tell
> > before someone seriously starts integrating medusa. Until then, the
> > medusa code doesn't hurt anyone, nor has it caused any maintainer
> > burden, so removing it sounds quite unnecessary.
> 
> O.K., so may I please remove all search-related files from
> po/POTFILES.in, so that it doesn't cause any translator burden as well?

I think I'm the developer who most wants to see search work in Nautilus,
but I cannot think if a good reason to keep the existing search code in
Nautilus.  All of it will be replaced--the interface with a usable G2
implementation, the service with a local call to a metadata db, the
views with Nautilus extensions, and it must work as a search folder to
work spatially.  The only reason I can think to keep the code is that it
takes time to remove it.  If someone is interested in how the search
code worked, they must checkout Nautilus 1.x and Medusa 0.5.1.

I'm sure it is ok to from the po/POTFILES.in.  Those messages are
irrelevant to Nautilus 2.x and Medusa 6.x

-- 
__C U R T I S  C.  H O V E Y____________________
sinzui cox net
Guilty of stealing everything I am.