Well the quota size is up to the container, and not based on the implementation.
For all you know some containers already compress data before storing it to disk/db/whatever right? The method of storing and the size of the data they are willing to store is completely up to them. You could post to the container's individual forums who's size restrictions you find to restrictive. And/or you could post to the opensocial spec group with a proposal to define a pre-requisite for the min size, but over all the spec tends to be very flexible in what a container should and shouldn't do (which makes sense since there's so many different kinds, even supporting app data is 'optional'), so chances of that being accepted are probably slim, but I could be wrong of course :) -- Chris On Fri, Jan 9, 2009 at 4:30 PM, Citron, David <[email protected]> wrote: > So am I the only developer who finds 10K a bit tight? :-) > > Thanks, > Dave > > -----Original Message----- > From: Citron, David [mailto:[email protected]] > Sent: Tuesday, January 06, 2009 11:46 AM > To: [email protected] > Subject: Compressing appData before persistence -- 10K is not enough > > Hi! Given that many containers (e.g. iGoogle) only allow 10K of > persistent storage per Gadget*User, and given that 10K is not > necessarily a huge amount of space (depending on your gadget), I was > wondering if there was any proposed strategy to compress persistent data > prior to storage. > > Client-side data compression is possible, of course, but there are no > standard JavaScript compression libraries that I could find, leaving the > options to: > - write my own JavaScript implementation of DEFLATE > - use an embedded Java applet to do DEFALTE (heavy!!) > - ? > > Server-side compression via zlib is easy enough with a custom Guice > module, but when a gadget is hosted on iGoogle/Orkut, etc. then the > server implementation is out of the control of the gadget implementer. > > What's the current recommendation? "Persist less"? :-) > > Are there plans to include some kind of data compression as a part of > the persistence infrastructure? > > Thanks! > Dave > >
