Re: Regarding GSoC 2012 project

On Sun, 2012-03-25 at 20:08 +0100, Dave Page wrote:
> On Sun, Mar 25, 2012 at 7:56 PM, mukesh agrawal <mkagrawal18(at)gmail(dot)com>wrote:
>
> > Hi Dave
> >
> > Actually I was having my mid-sem exams, so I could not much related to the
> > project, But in the meantime, I went through
> > http://wiki.postgresql.org/wiki/PgAdmin_Internals but I didn't get much
> > related to the project.
> >
> > In the project proposal, I had to answer about the project schedule(How
> > long will the project take? When can you begin work?) but I am still not
> > much clear about this question. Could you please help me in answering this
> > question?
> >
>
> I haven't investigated what work would be needed to complete those
> projects, nor do I have any idea of your capabilities so I'm not able to
> suggest a project schedule for you. You'll need to look at the code, and
> try to get an understanding of the requirements of each sub-project, and
> then estimate how long you think it would take you to complete them. I
> would then suggest doubling that estimate to allow for overruns and testing
> time. You should also bear in mind you'll need a week or so to get a build
> environment setup and become familiar with it, and that you'll begin the
> project with little experience of the code base and wxWidgets, so plan on a
> slow start to begin with. I will say that I think a competent C++ developer
> should be able to produce a polished, release quality result for this
> project in the time allotted for GSoC.
>
>
> > Could you also please suggest me about what I should do now related to the
> > project to strengthen my GSoC application?
> >
> >
> Research as mentioned above - and avoid under-estimating the requirements
> of the task (that's a common mistake, not just in GSoC, but in development
> in general). Show that you understand the project in the application, by
> including initial thoughts on how you think the code should be written
> (what are the important functions you'll have to modify, what will you need
> a parser for, how will the new features look to the end user etc)?
>

I totally agree with Dave here, especially on doubling the estimate of
time. Don't take it as a figure of speech. You really have to double
your estimate. Even with this, you may overrun.

--
Guillaume
http://blog.guillaume.lelarge.info
http://www.dalibo.com