Re: [Xen-devel] Xen Code Review Dashboard - Outreachy Program Project

["Jesus M. Gonzalez-Barahona" <jgb@xxxxxxxxxxxx>]

On Tue, 2016-10-04 at 20:18 -0400, tevin.k.mallory@xxxxxxxxx wrote:
> Hello Jesus!
>  
> Thank you for taking the time out of your busy schedule to responded.
> I would love a summary of the small contribution over email, as this
> will allow me to get started on the project sooner. I am located in
> Florida USA, on Eastern Daylight Time and available anytime on
> October 6th to discuss details. If you have certain times that work
> best for you I can easily adjust my schedule. Just let me know when
> you would like to chat via IRC and I will be there. For the most part
> my schedule is very flexible, I am always available on Mondays,
> Tuesdays and Thursdays at anytime. Thank you once again and I look
> forward to hearing from you.

OK, here we are. With respect to the project in general, I assume
you're familiar with

https://wiki.xenproject.org/wiki/Outreach_Program_Projects#Xen_Code_Rev
iew_Dashboard

Right?

The main aim of the project is to reproduce all the process up to
having something like https://xen.biterg.io/ But using only GrimoireLab
tools

http://grimoirelab.github.io

Now it is using a mixture of MetricsGrimoire and custom scripts

The first step is to get information both from mailing lists and git
repositories using Perceval, and storing it in ElasticSearch. Later,
there are some scripts that should be ported to use this ElasticSearch
data (instead of the SQL data they are using now). With that, produce
the ElasticSearch indexes for the dashboard. Then, if possible, improve
the dashbosrd and make it more useful for the Xen community

Now, about the microtask.

I guess you know you can start with one microtask to show that you are
likely to be the right person for this project, according to Outreachy
requirements.

In this case, the microtask would be getting data from a mailing list
with Perceval. The mailing list is xen-devel. You get its archives
analyzed by Perceval, and store the resulting raw index in
ElasticSearch.

Once you're good with this, you do the same for the Xen git repo

And once you're good with this, you write an script which produces a
new index with some information for each commit, plus the branch in
which it was committed

Mboxes are at

https://lists.xenproject.org/archives/html/mbox/

They are then named xen-devel-<year>-<month>, e.g. xen-devel-2016-09

You can write a little script using wget or curl to downoad several of
them at once. To begin with, you can start with some of them (say 5-10)

The code contribution result of the microtask would be the
identification of the branches, based on the output of Perceval / git

The setup would be you getting all the info from some mboxes in
ElasticSearch, a git repo in ElasticSearch, and a simple index,
combined of both plus branches information, again in ElasticSearch

Once you have those in Elasticsearch, just produce the result of
querying some of the items in ElasticSearch with curl, and the code for
the identification of branches. All of this can be stored in a git
repository for verification.

I can support you via irc and email if you have any trouble.

We're compiling some information on how to use GrimoireLib in

https://jgbarah.gitbooks.io/grimoirelab-training

Maybe that's a good place to start.

Is all of this ok with you?

        Jesus.

-- 
Bitergia: http://bitergia.com
/me at Twitter: https://twitter.com/jgbarah


_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxx
https://lists.xen.org/xen-devel