[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index] [List Home]
Re: [p2-dev] Use of uncompressed metadata repositories

Most people should be using the zipped format.  I'm personally not too fussed about the uncompressed version, as we have a solution (compress it).  There is at least one good use for the uncompressed version however, and that's if you are only accessing local repositories (and you have lots of disk space). In this case, the extra time to unzip the repo might not be worth it.

There is another interesting problem with our repositories (and the duplication of licenses), and that's the amount of garbage we produce when parsing. While we do use a String pool, we first create separate strings, then find the duplicates in the pool.  I noticed that parsing a 50M (unzipped) repository took over 200M of heap space.


On Wed, Nov 17, 2010 at 10:22 AM, Paul Webster <pwebster@xxxxxxxxxxxxxxxxxxx> wrote:
On Wed, Nov 17, 2010 at 10:00 AM, Dean Roberts <Dean_Roberts@xxxxxxxxxx> wrote:

So does anybody have an opinion on how widely used uncompressed repositories are?

I can't comment on other projects, but we took the defaults for PDE build in e4 and it looks like we get a content.xml and artifacts.xml, instead of the jar.


Paul Webster
Hi floor.  Make me a sammich! - GIR

p2-dev mailing list

R. Ian Bull | EclipseSource Victoria | +1 250 477 7484
http://eclipsesource.com | http://twitter.com/eclipsesource