streams-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From sblackmon <sblack...@apache.org>
Subject Streams + Juneau Microservice Example
Date Fri, 21 Apr 2017 16:40:17 GMT
I've got a pretty interesting micro-app based on juneau and streams coming together.  

Its an html-enabled proxy for accessing profiles, posts, and connections across multiple social
networks, using a common API and data format based on the W3C SWG ActivityPub and ActivityStreams
specs.

The consolidation we’ve started around Juneau as the principle library for interacting with
data providers made this module (and its dependency tree) much more straight-forward and simple
than it would have been prior.

I think some version of this could ready to release / demo / blog by ApacheCon. If anyone
wants to test it out and contribute feedback or features, that would be excellent.

You’ll need to checkout and build the steveblackmon/STREAMS-504 branches of incubator-streams
and incubator-streams-examples, then read incubator-streams-examples/servers/streams-server-juneau/README.md
which will explain what to do next.

Cheers,
Steve
On January 14, 2017 at 10:28:49 AM, Steve Blackmon (sblackmon@apache.org) wrote:

Good idea. I can work on wrapping several streams data collection, conversion, and storage
modules with Juneau microservice.  We were planning to publish a dropwizard example with
our next release anyway, let's use Juneau instead.  Should make for a good ready to run demo
and blog post for sure.

On Jan 14, 2017 8:49 AM, "John D. Ament" <johndament@apache.org> wrote:
Just wondering, do we want to put out a blog post on the conversion?  This
is a great demonstration of synergy between podlings.

Thanks for picking this up Steve!

John

On Fri, Jan 13, 2017 at 4:43 PM sblackmon <sblackmon@apache.org> wrote:

> James,
>
> I was able to switch several APIs from dropwizard to juneau-microservice
> in an afternoon, removing jackson in the process.  Lines of code decreased,
> configuration scheme was simplified, and got HTML and XML support for
> free.  Very nice.
>
> I suspect that JSON-LD operations would need to use RDF libraries for
> anything beyond the simplest examples.  Jena provides some support for
> JSON-LD already. [1]
>
> [1] https://jena.apache.org/documentation/io/
>
> Steve
> On January 6, 2017 at 8:32:18 AM, James Bognar (
> james.bognar@salesforce.com) wrote:
>
> Hi Steve!
>
> I welcome your interest in our project! We're looking for ways to increase
> external participation.
>
> Juneau's been a pet-project of mine for a while as an internal IBM
> community source project. There are probably differences between Jackson
> and Juneau since I specifically did NOT look at the Jackson implementation
> so that it would be a cleanroom implementation. However, I know teams that
> successfully converted from Jackson to Juneau. One team did performance
> tests and showed that Juneau was 20% faster than Jackson for a base set of
> data and that the performance improved with larger datasets.
>
> I haven't looked at JSON-LD specifically. I'll have to investigate to look
> at the feasibility. One possible issue I see is that Juneau doesn't use a
> DOM for serializing/parsing for any of the languages (except the Jena-based
> RDF ones). That's one of the reasons why it's efficient....it marshalls
> directly from POJOs.
>
> I'll join the mailing list and contribute any ideas I have.
>
> Thanks!
> James.
>
>
> On Thu, Jan 5, 2017 at 12:29 PM, sblackmon <sblackmon@apache.org> wrote:
>
> > Hello,
> >
> > Juneau came to my attention recently and I’ve just had the chance to read
> > through the documentation and play with the micro service samples.
> > Congrats on your second apache release.
> >
> > I’m on the PPMC of Podling Streams - http://streams.incubator.apache.org
> ,
> > Your project is interesting to me and I just wanted to say hello and
> start
> > a dialog.
> >
> > We use jackson and typesafe config heavily - so there’s a possibility we
> > could switch to juneau for those capabilities as it appears you already
> > have feature and (potentially?) performance parity.
> >
> > We maintain hundreds of JSON schemas from which we generate POJOs which
> > are used within the project. The code-gen process supports internal and
> > external $ref - I noticed that Juneau has similar code so perhaps we
> could
> > combine efforts on that front.
> >
> > In addition to POJOs we’ve got maven plugins that can use json schemas to
> > generate scala, and resource files to configure platforms including
> > cassandra, elasticsearch, hive, and pig to store / analyze compliant json
> > documents. This is a relatively immature capability that we built
> > ourselves because a) maintaining files like that is a headache and b) we
> > didn’t see anyone else addressing it using schemas.
> >
> > One specific question I have about Juneau is whether you plan to add
> > support for JSON-LD in 2017. We are putting together a plan to support
> > Activity Streams 2.0 and will need a fairly sophisticated JSON-LD engine
> to
> > make that work.
> >
> > I welcome anyone from your community to join our list and suggest other
> > ways we might collaborate.
> >
> >
>
>
> --
> James Bognar
>

Mime
  • Unnamed multipart/alternative (inline, None, 0 bytes)
View raw message