[storm] What about MondoDB?

[Eduardo Willians]

As all of you should know there's MongoDB out there. I loved it when I
came out to know it. I really think Mongo will become very popular and
its importance wiil rise more and more, IMHO.

I was looking some ORMs for mongo (on top of pymongo), and there are
basically two: MongoKit and MongoEngine. MongoKit does not look a
really ORM, and MongoEngine was built on a django-model style, which I
think is not good. There's a lot of magic, and things like:

cls_spec.objects(field__lte=32)

Instead of our good python-with-not-too-magic formula:

store.find(ClassSpec, ClassSpec.field <= 32)

Then I decided to make some changes in MongoEngine to try to make it
work just like Storm, I mean, separating ClassSpec from Database and
create a powerful ResultSet class and reducing inheritance programming
way.

Well, today, when I was studying MongoEngine's code and stuff, I had
another idea: why not creating another backend-layer to allow Storm to
mappering Mongo?

I know Mongo is not a relational database, and there's the fact that
Mongo allows direct inheritance and that allow embedeed document, and
allow relational fields, but, on front-end, to mapper it won't be so
diferent of what Storm front-end does.

Good. You may say: "open source code is out there to be contributed
to, so contribute to it". Yes. And here is my proposal, I can try, but
I'll need help. And I won't try it unless storm-team allow me to try
it, since I'm not a powerful programmer I'm not able try it alone, but
here I am.

Actually, I would like to already have it implemented, for me it's
more important to have the product than to produce it. So, if any geek
take the leadership of such possible branch, it will be a lot easier,
but as I said, I'm here.

from brazil.campo_grande import ThankYou

Eduardo Willians