Benoit's interview

[Akshat Jiwan Sharma <ak...@gmail.com>]

Hello everyone,

Benoit just posted the answers of the interview questions to me.  This
could go on this weeks of couchdb news or should it go in a seperate topic?
Not too sure about it. What do you think?

I am copy pasting this from the email:-


0. Tell me something about yourself. How long have you been working on
couchdb? What excites you about it?

1. What was the motivation behind the rcouch fork?

The initial reason was to create a pure Erlang release that could be easily
customized for specific needs. Especially for my customers and the refuge
project. It was also a way for us to move fast and freely on new features.



> 2. How did the structuring of rcouch differ from couchdb?
>

With the new master, the organization is quite the same even if we pushed
further the split between applications like the http api and some others
applications. We also have the possibility to build a static release

>
> 3. What challenges you faced/are facing in merging rcouch with couchdb?
>

The main difficulties is to keep the history of the couchdb project while
maintaining our own set of patches. Since the pace of updates in couchdb is
not that fast, updating rcouch doesn't require too much work though.



>
> 4. How is the "_changes for view" different from _changes with a _view
> filter?
>
>
The view filter that I added in couchdb is just wrapping the view function
and pass it over all docs in the database like any filter. It can be really
slow to do it when you start from 0.

The view changes however is done directly on indexes without evaluating the
JS function each times. It's much faster and efficient. It also offers the
possibility to merge different views and listen on events. More information
can be found on that page:

https://github.com/rcouch/rcouch/wiki/View-Changes



> 5. Does the rcouch play well with the big couch merge?
>

Merging the view changes and other features should be easy. Merge is in
progress.


>
> 6. Cheesecakes or chocolate pudding?
>

None really.  I prefer more simple sweets with less fat.


>
> 7. Will "_changes for view" work with third party extensions like
> couchdb-lucene?
>

I don't know. It can probably work since you're able to replicate docs
using a view change.


> 8.  What is the refuge project?
>

It aims to build a social data platform. Ie. It will define the protocols
needed to connects data sources between machines and people, coming from
devices, peoples or services in a decentralized manner. An implementation
of these protocols is also provided written in Erlang and other languages.

One simple example of what will be possible with the refuge platform, is
building a network of connected nodes on which you will be able to exchange
some blobs independently of the storage service. You will be able to
exchange smart contract to define the access and usages of a blob or a
collection of blobs in a P2P manner.

Re: Benoit's interview

[Andy Wenk <an...@apache.org>]

Hi Akshat,

first of all thank you very much for doing the interview. I am really
looking forward for other interviews as well.

I think this is very raw now. In my opinion, this should now be put into an
article form with some introduction concerning the motivation to do these
interviews. If the text is ready I am happy to review it. But we should
make sure that Benoit will acknowledge the article and then a native
english speaking person should copy edit it as well.

What do you think about this process? What do the others think?

I would give the article a "CouchDB Interview series No. 1 - Benoit
Chesneau" title or sth. like this. The Blog is a good place to post it then.

Thanks and Cheers

Andy



On 29 July 2014 18:42, Akshat Jiwan Sharma <ak...@gmail.com> wrote:

> Hello everyone,
>
> Benoit just posted the answers of the interview questions to me.  This
> could go on this weeks of couchdb news or should it go in a seperate topic?
> Not too sure about it. What do you think?
>
> I am copy pasting this from the email:-
>
>
> 0. Tell me something about yourself. How long have you been working on
> couchdb? What excites you about it?
>
> 1. What was the motivation behind the rcouch fork?
>
> The initial reason was to create a pure Erlang release that could be easily
> customized for specific needs. Especially for my customers and the refuge
> project. It was also a way for us to move fast and freely on new features.
>
>
>
> > 2. How did the structuring of rcouch differ from couchdb?
> >
>
> With the new master, the organization is quite the same even if we pushed
> further the split between applications like the http api and some others
> applications. We also have the possibility to build a static release
>
> >
> > 3. What challenges you faced/are facing in merging rcouch with couchdb?
> >
>
> The main difficulties is to keep the history of the couchdb project while
> maintaining our own set of patches. Since the pace of updates in couchdb is
> not that fast, updating rcouch doesn't require too much work though.
>
>
>
> >
> > 4. How is the "_changes for view" different from _changes with a _view
> > filter?
> >
> >
> The view filter that I added in couchdb is just wrapping the view function
> and pass it over all docs in the database like any filter. It can be really
> slow to do it when you start from 0.
>
> The view changes however is done directly on indexes without evaluating the
> JS function each times. It's much faster and efficient. It also offers the
> possibility to merge different views and listen on events. More information
> can be found on that page:
>
> https://github.com/rcouch/rcouch/wiki/View-Changes
>
>
>
> > 5. Does the rcouch play well with the big couch merge?
> >
>
> Merging the view changes and other features should be easy. Merge is in
> progress.
>
>
> >
> > 6. Cheesecakes or chocolate pudding?
> >
>
> None really.  I prefer more simple sweets with less fat.
>
>
> >
> > 7. Will "_changes for view" work with third party extensions like
> > couchdb-lucene?
> >
>
> I don't know. It can probably work since you're able to replicate docs
> using a view change.
>
>
> > 8.  What is the refuge project?
> >
>
> It aims to build a social data platform. Ie. It will define the protocols
> needed to connects data sources between machines and people, coming from
> devices, peoples or services in a decentralized manner. An implementation
> of these protocols is also provided written in Erlang and other languages.
>
> One simple example of what will be possible with the refuge platform, is
> building a network of connected nodes on which you will be able to exchange
> some blobs independently of the storage service. You will be able to
> exchange smart contract to define the access and usages of a blob or a
> collection of blobs in a P2P manner.
>



-- 
Andy Wenk
Hamburg - Germany
RockIt!

GPG fingerprint: C044 8322 9E12 1483 4FEC 9452 B65D 6BE3 9ED3 9588

 https://people.apache.org/keys/committer/andywenk.asc