[kepler-dev] Kepler distribution?

[Matt Jones]

Hi Tobin,

Yeah, getting nightly builds to be available was exactly how this thread 
got started.  I proposed in that earier message that we make the 
products of the nightly Anthill build available, and Dan suggested the 
JWS format.  I think this would be a good idea.

The testdata is unfortunate but needed for the time being.  The Garp 
will eventually be removed, but for now we need it to test whether 
changes break existing geospatial actors.  Over the medium term, our 
data access components that we are building will be able to access this 
data from the grid and cache it locally after its been downloaded once 
-- so it will no longer need to be part of the build.

Matt

Tobin Fricke wrote:
> Re: The current talk of installers, etc -- Currently the only way (AFAIK)
> to get Kepler is to check it out of CVS.  I have made CVS snapshots for my
> colleagues who want to play with Kepler:
> 
> 	cvs checkout kepler
>         setenv archive kepler-`date +"%G%m%d"`.tar
> 	tar -cvf $archive kepler
> 	gzip -9 $archive
> 	mv ${archive}.gz ../public_html/archive
> 	ln -sf ../public_html/{${archive}.gz,kepler-current.tar.gz}
> 
> Could we make such snapshots available from the ecoinformatics site,
> generated by a cron job that runs periodically? (Even weekly would be
> fine.)  Anonymous CVS is another option (more complicated, but more
> efficient).
> 
> The Kepler CVS snapshot clocks in at 156 Mb uncompressed or 37 Mb gzipped
> -- a rather huge distribution!  "du kepler | sort -n" shows that this is
> heavily lumped in the 'testdata' (104 Mb) and, specifically,
> 'testdata/garp' (95 Mb) subdirectories:
> 
> ...
> 8616    kepler/lib/testdata/grass
> 20355   kepler/docs
> 23072   kepler/lib/jar
> 95215   kepler/lib/testdata/garp
> 104307  kepler/lib/testdata
> 130860  kepler/lib
> 159325  kepler
> 
> Removing testdata/garp results in tarball that is 63 Mb uncompressed and
> 34 Mb gzipped.   I guess that test data isn't too entropic. (-:
> 
> I think that a tarball distribution for linux is sufficient for (and
> preffered by!) most users of that platform, although some kind of
> demand-loading Web Start system would also be cool.
> 
> Tobin
> 
> _______________________________________________
> kepler-dev mailing list
> kepler-dev at ecoinformatics.org
> http://www.ecoinformatics.org/mailman/listinfo/kepler-dev

-- 
-------------------------------------------------------------------
Matt Jones                                     jones at nceas.ucsb.edu
http://www.nceas.ucsb.edu/    Fax: 425-920-2439    Ph: 907-789-0496
National Center for Ecological Analysis and Synthesis (NCEAS)
University of California Santa Barbara
Interested in ecological informatics? http://www.ecoinformatics.org
-------------------------------------------------------------------