Emmanuel,
Thanks for your reply!
Unfortunately -- perhaps this was not clear to you -- we really want the
application to be *the student's work*, to ensure that the student has
come up with the implementation ideas, can communicate clearly, can take
feedback, and so on. Please see
http://www.booki.cc/gsoc-mentoring/defining-a-project/ and
http://www.booki.cc/gsoc-mentoring/selecting-a-student/ especially
"starting at the beginning" at the end of "Selecting a Student".
--
Sumana Harihareswara
Engineering Community Manager
Wikimedia Foundation
On 05/02/2013 03:36 PM, Emmanuel Engelhart wrote:
Dear Kiran, Dear Sumana
I have worked on the application. I have rebased it on the WMF template
like suggested by Sumana.
@ Kiran
I think it's really important to complete/pimpup the last paragraphs
which are about you, your motivations, past experiences: people should
be convinced about the project *and* about you.
@Sumana
Thank you for your valuable feedbacks. My remarks:
* I have add portability constraints. We have a compilation farm and VM
to tests.
* With average Wikipedia ZIM file, the incremental update should be able
to save ~70/80& of the bandwidth.
Kind regards
Emmanuel
Le 02/05/2013 13:45, Sumana Harihareswara a écrit :
On 05/01/2013 07:03 PM, Kiran Mathew Koshy
wrote:
Hello,
I have submitted my project application for GSoC '13. Please review it.
Link:
https://www.mediawiki.org/wiki/User:Kiran_mathew_1993
Thanks,
Hi and thanks for your interest in improving MediaWiki!
https://www.mediawiki.org/wiki/Mentorship_programs/Application_template
would be a very nice thing to follow so we have the information we need
to evaluate your application. I especially encourage you to link to
your past code and open source contributions.
Some questions you ought to answer:
* What *kind* of documentation will you be writing? A walkthrough of
the code so future developers can understand it? Some kind of
user-facing manual? Other things?
* Will you be incorporating any kind of automated testing into this
project? To guarantee the robustness of the proposed process, it's far
better to have self-checks of some kind.
* What are your plans for testing this on Windows, Linux, and Mac OS?
Do you have test environments available for all those operating systems?
* Have you done any measuring to figure out how much bandwidth this
would save, given usual usage patterns?
Thanks,
Sumana