Is there a plan to add a CONTRIBUTING.md? This could include template to do a PR, guidelines to add tests, coding standards, etc.

I'm currently learning Rust and am very interested in contributing on this project somehow. Thanks!

When I add the document with same id but different content to the index, the former still can be searched.
Maybe it should be replaced by the latter.

How can I set the name of the top/root directory so that its not the domain name?
Potentially a future feature to set a custom name?

Description

We need better guidance around choosing hardware. For a given size and shape of data, how much memory do you need? How much CPU should you provision to hit certain latency targets, how much disk and how fast, etc.

This won't be easy, and will need to be supported by benchmarks.

Related:
#1796
This will be helped by #1926 and improving our statsd stuff and usage of it.

Now, some query examples(simple and prefix) are given, but it seems to be not enough to get quick start for more complicated query logic.

Query String Query supports all kinds of query such as Phrases, Field Scoping, Boolean Query, Numeric Ranges, fuzzy search and so on.

If more examples are given, it seems to be more friendly to tiro.

But as we are currently targeting JDK 8, and a new API arrived in JDK 9, it does not make sense to do this yet. The next long-term Java version 11 is scheduled for end of September 2018.
So for ELKI 0.8 it is an option to target JDK 11, and use the new API then.

git apply PostgresNode.patch.txt
echo 'exec valgrind postgres "$@"' > /tmp/pgvalgrind && chmod +x /tmp/pgvalgrind
PGLAUNCHER=/tmp/pgvalgrind make check -C contrib/rum

in command CREATE INDEX rumidx ON tst USING rum (t rum_tsvector_ops); will be endless working. because calling the function repalloc calls the macro VALGRIND_MEMPOOL_CHANGED which is very slooooow.

I suggest r

In case someone wants to add extra content (e.g. an additional script ) in the existing template, it would be good to be able to do it also from inside template function (instead of copying/editing the default template). To achieve this, the default template should be added in locals object passed to template function.

The advantage of this is that on a later package upgrade, the template wil

Hi, i am currently using your services to implement an instant search to improve system search. And everything looks great till now. But currently i m getting an issues with a model import data, since am using aggregator to use multiples models in one indice.

The issues is like so:

1. When i add some models, lets say, Contrato, Pagamento and User to the models array property of the Aggregat

I would like to be able to search multiple indices. I suggest the search function is extended so that the index parameter can be an array.

Rather than using the include/exclude options someone may want to identify files that should/shouldn't be included by marking them with a header at the top of the file. This is along the lines of tools like istanbul and tslint that allow checks to be disabled using a comment.

Should this be a predetermined string? I am not sure of any use cases where someone would need to use their own format.

The documentation shows no way to log into a solr server secured with user/password authentication.
Can it be done with a command line option?

This is a tracking issue for additional Indexer implementations.

Here is a non-exhaustive list of potential candidates for an Indexer implementation:

• RemoteIndex:
  Requires to implement a companion server.
  The idea is to have a clone of the index on another machine on which a companion server is running, you would then point Alexandrie to that server and it would take care o

The depends-on: query (such as accessed via "Artifact Dependencies") takes a parameter of form "$organization/$artifactName". The dependents and dependency boxes on artifact pages also display this form.

This is not consistent with the hierarchy a user sees elsewhere in Scaladex which has the general form "$organization/$project/$artifactName/$version", as the $project level is omitted.