[Clipart] Open Clip Art Library 0.07 Release

[Alan Horkan]

On Mon, 4 Oct 2004, Bryce Harrington wrote:

> Date: Mon, 4 Oct 2004 10:44:46 -0700 (PDT)
> From: Bryce Harrington <bryce at bryceharrington.com>
> To: Nicu Buculei <nicu at apsro.com>
> Cc: clipart at freedesktop.org
> Subject: Re: [Clipart] Open Clip Art Library 0.07 Release
>
> On Mon, 4 Oct 2004, Nicu Buculei wrote:
> > Bryce Harrington wrote:
> > > I uploaded a new version of the package with changes.
> > > Please update the numbers on the website.
> >
> > this version is better: down to 2139 files (only 385 duplicate files
> > left) and 13.8M (only 3.9M duplicated of zip archive)
>
> It sounds like you don't understand that there will be duplicates.
> Since each keyword corresponds to a subdirectory, for each file that has
> multiple keywords, there will be multiple copies.  Please see the thread
> from a couple weeks ago for the need for this approach plus the longer
> term goals.

I thought the keywords would be used to avoid the need to generate
directories and that a browser/search tools would be built on top of it or
that applications would be expected to have their own systems of managing
and sorting through lots of extra add-ons.

But maybe that will happen in future! (I'm being optimistic).

For an application like Inkscape to manage the huge amount of clipart
available from OpenClipart.org it would need to write a basic File Manager
like interface to do it properly.  It is unfortunate that Inkscape the
biggest potential user of OpenClipArt.org does not really have a suitable
system for managing large amounts of files.

Perhaps instead a developer might be interested to develop a more generic
"Clipart-App" that unlike Nautilus would be portable and although
resembling a file manager it would provide a more abstract view of
collections/albums rather than a literal filesystem view.

If it were still around the GUASH file browser that Gimp used to have
might have been a good basis for an application like this.

> > >>i was intrigued by the huge increase, from 1700 files (reported supposed
> > >>to be in previous release) to 2300 files (supposed to be now) so i
> > >>counted the unique files in this release: the correct number is *1754*.
> > >>so we have about 600 duplicate images (1800 total files, including .png
> > >>and .txt)

The reported 7 MB size difference is worrying, but I suspect that 7MB is
the size on disk and that the duplicates would have compressed reasonably
well and not made a significant difference to the total size of the
compressed archive.

- Alan