On Sat, Jan 4, 2014 at 5:08 PM, Hans J Schroeder <[email protected]> wrote:
> I am glad that you are aware of the feature and its advantages. The prefix
> solution is a clever idea. It would certainly help, but I think it is to
> restricted for the choice of IDs. Wouldn't it be possible to use an
> existing type index as a pre filter? I guess every document should have a
> type attribute and the database should have a type index.

Why is the prefix ID too restrictive? Do you have concrete use cases
that would be prohibited by this kind of implementation? The problem
with using a document attribute is having to parse the entire document
first. By using the document ID directly, you can decide whether or
not to parse and process the document based on just the ID, which
would make sense if one of the goals is to make things faster.

> Anyway, is the prefix solution on the release plan?

We don't have that kind of release plan. We generally keep track of
feature proposals through JIRA issue. It might well be we already have
an issue for this, if you want to search for a bit and if you can't
find anything open one, that might be useful. After that, someone
would have to have the time and inclination to actually implement it
(if you don't have anyone like that in your organization, putting out
a bounty may help).

Cheers,

Dirkjan

Reply via email to