Discussion on GSoC: appengine support
nasim
I am very new to django. I have built some web apps with django which
are hosted in google appengine. To built this, I had to use appengine-
patch and django-appengine-helper. I also needed to do some of my own
patches into the django code. I really want a fully appengine
supported django so that I don't have to patch it here and there and
cripple site and django's functionality.
I am very much interested to start working on solving this problem.
Although mentioned in django-gsoc-wiki page, it's a huge task, I want
to start somewhere. Schemaless db support may be the entry point. I
haven't read very much django itself. But I am strongly willing to do
that. I also do not have any opensource development experience yet.
I want to apply for google's summer of code this year through this
project: implementing schemaless db in django. Please show me some
directions.
Thanks.
Russell Keith-Magee
Hi Nasim
Before I begin, let me preface this by saying that I really am a nice
person, and I'm not trying to deliberately crush your spirit. Please
don't take these comments personally - I'm just trying to let you know
why your proposal probably won't be successful in its current form.
The GSoC selection process requires two things:
1) You need to want to do the project
2) We need to believe that you're capable of completing the project
I have no doubt you meet the first criterion. However, you need to
convince us of the second point, and this proposal is a long way from
achieving that goal.
The wiki page says this is a huge task for a very good reason - it's a
_huge_ task. GSoC has something like a 12 week work schedule, in which
you need to design, build, document and test an entirely new database
backend, using a completely different paradigm for representing data.
On such a tight schedule, there isn't going to be a lot of time for
coming up to speed - with anything. If we were to select an AppEngine
project for GSoC, the application would need to have an existing
demonstrated knowledge of:
1) How the internals of app engine work
2) How the internals of Django work
3) The underlying philosophy of a column-based data store like AppEngine
4) The underlying philosophy of Django's ORM
5) How a column-based data store would map onto Django's ORM
Based on the information you have provided so far, it's not clear that
you meet any of these prerequisites. In fact, you announce that you
are a novice to Django, and haven't worked on an Open Source projects
before - so
in addition to any technical knowledge that you are missing, you may
also need to learn how the open source development process works, and
how Django's development process works in particular.
I'm not saying that "novices need not apply". We would be only too
happy for a novice to apply and be accepted to the SoC (and we have
accepted novices in the past). However, if a novice was to be
accepted, they will need to prove to us that they have an idea of what
is required, how much effort is required, and the standard to which we
expect work to be completed. A novice also needs to be careful to
balance their expectations against their novice status - unknown
novices proposing to major architectural changes aren't going to be
taken seriously without some serious work.
Students with whom we have an existing relationship with the Django
project have an obvious advantage here - but the door isn't closed to
newcomers. It wouldn't be hard to convince us that you have what it
takes. The first step is to demonstrate that you can work
independently by coming up with a clear and concrete proposal. It
doesn't need to be a final proposal, but it does need to be rich
enough to prove to us that you have at least thought about the problem
you are proposing to solve at a level deeper than "Hey, that sounds
like fun!".
If you want to see how to do this properly, go back 1 year in the
mailing list archives (django-dev, I think), and look for messages by
Nicolas Lara about doing Aggregation for the GSoC. I worked with
Nicolas last year, and his approach to his GSoC application was a
model for all applicants to follow.
In short, we're not going to "show you directions". I can guarantee
that we will receive more applications for GSoC positions than we have
available positions. If you want one of the limited slots that are
available, you have to prove to us that you are worth it. We will be
only too happy to help finesse a proposal, or point out potential
problems, or discuss design suggestions, but we won't build your
proposal from scratch.
Yours,
Russ Magee %-)
nasim
look into the previous applications and hopefully come up with some
idea. I am not giving up any hope yet. Wish me luck.
Yours,
Nasim