On Jun 13, 2007, at 4:16 PM, Niklas Lindström wrote:
Hello!
I think a separate package would be good. To keep RDFLib as a core
package seems fine; this tool package could then become more rich than
would suite a core package, e.g. a WSGI SPARQL entry point,
filesystem-to-graph-tools etc. (Both of which I have some code; and I
believe e.g. Chimezie also has a SPARQL-endpoint (CherryPy-based).).
My thinking is along these lines as well -- think it'd be good to
keep rdflib a core library. So we'll keep an eye on when these tools
should take on a life of their own.
I believe I mentioned thinking about this earlier this year. Still, I
think it would be great if you want to be its creator though, among
other reasons for the more "official air" of it. Still, is there a
reason not to have it in the same repo as RDFLib itself? It could for
instance ease any future move of some parts into the core; and of
course to allow the same users as for RDFLib itself. (Apart from that,
a google code repo would do fine too.)
Ah, mystery solved re: who else was thinking about starting such a
project. Think shared repo / overlap in developers where some of the
reasons why we didn't start a separate project for it from the start.
Some ideas to boot:
.. It is nice that setuptools provide the means to automatically add
command line tools (the 'console_scripts' entry point) - as you have
already set it up. Apart from the current conversion tool, we could
then add e.g. a sparql tool and some kind of validator.
.. For an example of what I mean with "filesystem-to-graph" stuff, see
<http://oort.googlecode.com/svn/trunk/oort/util/graphs.py>.
Neat :)
Best regards,
Niklas
On 6/13/07, Daniel Krech <[EMAIL PROTECTED]> wrote:
Hi all,
As of the last release we included an rdflib_tools top level
package to
start gathering some unreleased tools people had around. This was
meant to
be a temporary place holder until a separate project could be
created. Any
reason not to split these off as a separate project as planned?
Main reason
motivating the split now is an issue that was hit with the debian
packaging.
I remember someone telling me they wanted to create such a
project, but
don't remember who. Anyone wanting to start and lead such a
project? Else
I'm willing to create a project at code.google.com that we can
move the
rdflib_tools bits to.
Daniel Krech, http://eikeon.com/
_______________________________________________
Dev mailing list
Dev@rdflib.net
http://rdflib.net/mailman/listinfo/dev
Daniel Krech, http://eikeon.com/
_______________________________________________
Dev mailing list
Dev@rdflib.net
http://rdflib.net/mailman/listinfo/dev
