[Mailman-Developers] GSoC 2014 : Proposal for the Mailman CLI project

[Rajeev S]

Hi Stephen,

The CLI project would be a sub module for the mailman.client project.

Since bzr does not have the submodule feature, I must be doing it either by
using a new repository or as a new branch to mailman.client .The latter
would be better as it would be easier to integrate the code into the
mailman.client project when this project is completed.

Now about the implementation part.As described in the project timeline, the
first phase would be to build the command tools. I would be building two
classes List and Domain, and identify the various methods that are to be
given to them. Many of the methods are identified in my GSoC proposal. The
command parsing would be handled by the python Optparse library.

I would like to start building the Version 0, but not to throw away, but
will be refining it further as per the feedback. If all this is OK I would
start building the version and push it to the mailman.client project.

And forget about the git vs bzr part. I am OK with using bzr :).




*Regards,Rajeev S*
*Government Engineering College,Thrissur*
*http://rajeevs.tk <http://rajeevs.tk>*


On Tue, Apr 22, 2014 at 9:01 PM, Stephen J. Turnbull <stephen at xemacs.org>wrote:

> Just to follow up quickly (I've got problems I need to deal with
> elsewhere over the next couple days).
>
> Abhilash Raj writes:
>
> Hey, thanks for jumping in, maxking!
>
>  > Hi Rajeev,
>  >
>  > Congratulations! We look forward to a great summer with you.
>
> Definitely!
>
>  >> I would like to thank the Mailman community and mentors for
>  >> their extensive > support during my application process,
>
> You're very welcome.
>
>  >> Also I have a few questions as part of the community bonding
>  >> process.
>  >>
>  >> 1.Is the project to be developed as an independent project or as a
>  >> part/branch of the Mailman Core repository
>  >
>  > That is supposed to be an implementation detail of your
>  > proposal. What do you think would be the best?
>
> I agree with Abhilash.  You make a proposal, we'll criticize it
> (contructively; "criticism" is not necessarily negative, just as with
> literary critics we may give a positive review -- you should know when
> you're doing things right).
>
>  >> 2.If it is an Independent project, Is it OK to use the >
>  >> git+gitorious/savannah or should I stick to bzr+launchpad? I have
>  >> used git extensively,naturally more comfortable with git.
>  >> However, I can pick up bzr if necessary.
>
>  > Even if it is an independent project we strongly encourage you to
>  > use bzr, as integration and code review might become a problem in
>  > later stages of your project( I am speaking from personal
>  > experience). Learning bzr and launchpad is not at all difficult if
>  > you know git.
>
> The CLI won't be *that* independent.  It really needs to be on
> Launchpad.
>
> Note: It may be possible to use git-bzr to fetch and push from
> Launchpad.
>
>  >> 3.Is it necessary for me to hangout in the IRC?If yes, when?
>  >
>  > It is not *necessary* for you to hangout on IRC. You just need to
>  > keep your mentors updated about what are you doing all the time. If
>  > you prefer email, I don't think that would be a problem. Your only
>  > responsibility is to deliver the project you proposed on time.
>
> It's generally a good idea to have interaction (brainstorming) when
> working on design.  I can generally be available on IRC from about
> 01:00 UTC to 15:00 UTC.  As Abhilash knows :-) I can often be
> convinced to stay around until about 17:00 UTC.  yaseppochi @ freenode
>
>  >> 4.Can I start coding right away?
>  >
>  > You can, but first be sure you know what you want to code. Your
>  > proposal does have a detailed description on working of the tool,
>  > however there is little mention about the details of implementation
>  > and design. It would be best if you first consult with your mentors
>  > and decide on something so that you don't waste time writing code
>  > that is not needed.
>
> Read Fred Brooks' /The Mythical Man-Month/, especially the essay
> "Build One to Throw Away".  Then think about whether you really
> believe him. :-)
>
> Regards and looking forward to working with you!
>
> Steve
>
>