incubator-blur-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Aaron McCurry <amccu...@gmail.com>
Subject Re: How do I present Blur
Date Fri, 06 Jan 2017 10:06:38 GMT
On Thu, Dec 29, 2016 at 4:22 PM, Mark Kerzner <mark@elephantscale.com>
wrote:

> Hi, Aaron,
>
> This all sounds like we are thinking along the same lines. Here are the
> things I could work on.
>
> 1. Upgrade Lucene. But I suspect I might ruin something, and you are a
> better candidate for this.
>

This task is tricky because of several areas where the Lucene project makes
regular internal changes (protected Lucene APIs) and when they change I
usually have to re-implement some aspects of the project.  As with any task
I wouldn't discourage anyone from working on it because regardless if it
get totally completed or not whomever does the work always learns a great
deal about the technologies in the project.  We can just create a branch
and see how far we get.


>
> 2. Restore the UI submission.
>

As you saw in Chris's reply I'm not really the person to ask about that.


>
> 3. Or another task that can let me slowly get up to speed.
>

Unfortunately I can't think of any at this point and Jira is likely out of
date and needs to be cleaned up.

At this point the code is stable but the libraries are dated.  One library
that might not be hard as Lucene to upgrade is Thrift.  Blur uses 0.9.0 of
Thrift and the last release was 0.9.3 and the client code has to be
generated and integrated.  Unfortunately there are at least two aspects to
this task that make it fairly complicated.


   1. Thrift's source code was actually pulled into the Blur source and
   repackaged so that it could be used in the same classloader (e.i. same JVM
   process) as another version of Thrift without issue.
   2. The repackaged source had a few tweaks over the years (I think :-/ )
   to fix certain issues that came up.  I realize how bad this is from a
   development perspective, but it happened a couple of times.  So a review of
   the git history on that area of the code is in order to understand what
   changed so that after the upgrade whatever changes were needed could be
   reimplemented in a better way.

This task is very much not a starter task but I'm not aware any real
starter tasks.  Even updating supporting libraries like guava etc is a pain
because of the tight dependency on Hadoop.  So years ago I just accepted
that I would just have to live with the libraries that came with Hadoop
because it was proving to be very difficult to work around them.

Sorry I don't have a better story for jumping into the project.

Aaron




> Thank you
>
> Mark Kerzner, President, Elephant Scale <http://elephantscale.com/>
> Book a call with me here <http://www.meetme.so/markkerzner>
> Mobile: 713-724-2534, Skype: mark.kerzner1
>
>
> On Thu, Dec 29, 2016 at 3:14 PM, Aaron McCurry <amccurry@gmail.com> wrote:
>
> > On Sun, Dec 25, 2016 at 2:18 PM, Mark Kerzner <mark@elephantscale.com>
> > wrote:
> >
> > > Hi, Aaron,
> > >
> > >
> > > Here is what I think it needs.
> > >
> > > Update to latest Lucene. Why do you think it will be a major
> undertaking?
> > > Are you going to re-architect at the same time?
> > >
> >
> > If I were to do just perform the upgrade it might require some
> > re-architecting but I would hope it would be minimal.
> >
> >
> > >
> > > Create a front end, maybe a service in AWS also.
> > >
> >
> > Agreed.  There was a GUI contrib a long time ago, but I'm sure it needs
> > some work at this point.
> >
> >
> > >
> > > I can ask a friend who was considering using Blur, what stopped him. We
> > > will know what would people need.
> > >
> >
> > I'm always interested to know other people's reasoning.
> >
> >
> > >
> > > I am interesting. My colleague at http://elephantscale.com/ might be
> > > interested too. I have a related project, https://github.com/shmsoft/
> > > FreeEed,
> > > and https://github.com/shmsoft/FreeEedUI. It 'd be good, potentially,
> to
> > > add Blur as a back end, an alternative to Solr.
> > >
> >
> > Hmm, ok I will take a look.  Thanks.
> >
> >
> > >
> > > Update the 2011 copyright. It needs to look alive and be alive.
> > >
> >
> > Yeah.
> >
> >
> > >
> > > Personnel reasons that you mention are people, not your personal
> > problems,
> > > I assume.
> > >
> >
> > No I didn't have issues with anyone in the project or outside the project
> > for that matter.  My job position changed and the new position hasn't
> > required me to work on search related tech yet.  Along with the job
> change
> > I have had to spend far more time at home due to a health related issue
> in
> > my family.
> >
> > Thanks for the questions and the comments!
> >
> > Aaron
> >
> >
> > >
> > > Cheers,
> > > Mark
> > >
> > > Mark Kerzner, President, Elephant Scale <http://elephantscale.com/>
> > > Book a call with me here <http://www.meetme.so/markkerzner>
> > > Mobile: 713-724-2534, Skype: mark.kerzner1
> > >
> > >
> > > On Sun, Dec 25, 2016 at 11:30 AM, Aaron McCurry <amccurry@gmail.com>
> > > wrote:
> > >
> > > > It is, but without a community of active developers it has become
> > > > stagnant.  For example the Lucene library version it utilizes has
> > become
> > > > outdated and it would likely be a major undertaking to update the
> code
> > > base
> > > > to the newest version.  The biggest reason for the low activity it
> > that I
> > > > haven't had time to work on the project due to personnel reasons.
> > > >
> > > > In it's current state is it very stable even at very large index
> sizes
> > > > however the upfront development effort to use Blur is very high by
> > > > comparison to ElasticSearch or Solr.  I believe this was the primary
> > > reason
> > > > Blur never really caught on in the community.
> > > >
> > > > Aaron
> > > >
> > > > On Sun, Dec 25, 2016 at 12:14 PM, Mark Kerzner <
> mark@elephantscale.com
> > >
> > > > wrote:
> > > >
> > > > > But,
> > > > >
> > > > > Isn't Blur a new approach arguably better than SOLR and
> ElasticSearch
> > > for
> > > > > big sizes?
> > > > >
> > > > > Mark
> > > > >
> > > > > Mark Kerzner, President, Elephant Scale <http://elephantscale.com/
> >
> > > > > Book a call with me here <http://www.meetme.so/markkerzner>
> > > > > Mobile: 713-724-2534, Skype: mark.kerzner1
> > > > >
> > > > >
> > > > > On Sun, Dec 25, 2016 at 11:12 AM, Aaron McCurry <
> amccurry@gmail.com>
> > > > > wrote:
> > > > >
> > > > > > Development has slowed to just the occasional bug fix and mail
> list
> > > > > > activity.  That's about it at this point.
> > > > > >
> > > > > > Aaron
> > > > > >
> > > > > > On Tue, Dec 20, 2016 at 6:57 AM, Mark Kerzner <
> > > mark@elephantscale.com>
> > > > > > wrote:
> > > > > >
> > > > > > > Hi, guys,
> > > > > > >
> > > > > > > I teach SOLR, and I mention Blur as an alternative with
a very
> > > > > > interesting
> > > > > > > and different approach.
> > > > > > >
> > > > > > > What bothers me is somewhat low activity on the code. What
is
> the
> > > > > > > incubating status of Blur?
> > > > > > >
> > > > > > > Thank you
> > > > > > >
> > > > > > > Mark Kerzner, President, Elephant Scale <
> > http://elephantscale.com/
> > > >
> > > > > > > Book a call with me here <http://www.meetme.so/markkerzner>
> > > > > > > Mobile: 713-724-2534, Skype: mark.kerzner1
> > > > > > >
> > > > > >
> > > > >
> > > >
> > >
> >
>

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