paulruizendaal wrote:

> Hi all,
>
> highscalability has some interesting write-up about google app engine
> and its datastore. One blogger is quite disappointed with its
> limitations:
> http://catherinedevlin.blogspot.com/2008/09/bigtable-blues.html
> http://catherinedevlin.blogspot.com/search/label/geekeventaggregator
>
> Google has this post:
> http://googleappengine.blogspot.com/2009/02/back-to-future-for-data-
> storage.html#
>
> It makes a bold statement:
> "This isn't an accident -- when you build a system that can scale to
> the size that Bigtable can there's no way to do a general purpose
> join on data sets that size and still have them be performant."
>
> Now my question is whether this is true: is it mathematically
> impossible to do distributed joins in a generalized way?
>
> Any ideas anyone?
>

No. Google is just saying that since they're the smartest people in the
world and they don't know how to do it, it must be impossible. A given
their impoverished model of data, they're probably right.