On Mon, 2011-02-14 at 18:02 -0800, Jonathan Vanasco wrote:
> I spent some time today with Pyramid.  It was a bit of a rollercoaster
> -- at times I was very excited, at others very stressed.  i figured
> i'd share my thoughts here for the maintainers.

Thanks.

> 1. on setting up an app, there were no instructions on how to setup
> for a postgres 'url' in sqlalchemy ; in fact - there was nothing in
> there regarding this setup.  i pulled it out of an existing project;
> one could probably look at pylons 1.0 (not pyramid) docs.
> 
> 2. i assume this is part of the template designed to show how to do
> stuff... however creating the initial 'models' table, with a sample
> entry, is misleading and not clear.  i'd suggest changing it to
> "sample_model" or something that is very clearly not needed for the
> application.

I assume you did not find the SQLAlchemy-based tutorial here:
http://docs.pylonsproject.org/projects/pyramid/1.0/tutorials/wiki2/index.html


> 3. i haven't figured out yet if its possible to handle multiple db
> connections in pyramid.  in pylons i have "sqlalchemy.write.url" and
> "sqlalchemy.read.url" and handle multiple databases due to master/
> slave clustering.  in order to get widespread appeal, something like
> this needs to be very apparent and easy to do - Rails accomplishes
> this in a few lines of configs.

It's possible I'm sure.  Pyramid doesn't care really.  It's ignorant of
SQLAlchemy.  There should indeed be documentation that describes how to
do this though.

> 4. i'm not sold on the MV design.  personally, I think it was risky
> departing from MVC, which has become the de facto standard and super
> familiar for developers across the web.  django has had issues selling
> people on their MVT , and its just an uphill battle.  i'm sure it has
> many advantages, i'm just speaking from a 'marketing ' view.

We didn't depart from the design, just the acronym.

> 5. the default application template should really generate a multiple
> views & models structure.  most people i've worked with strive very
> hard to keep the entirety of their models/views from being a single
> file.

We chose to do that and document how to turn these into a package here:
http://docs.pylonsproject.org/projects/pyramid/1.0/narr/project.html#modifying-package-structure
 . We could have done it in reverse too but did not.

> 6. also on the migration point, it would be good to discuss sqlalchemy
> reflection.

Integration documentation is important, yes.  We don't have enough of
it.

> The code looks solid, and I understand the rationale behind a lot of
> design choices... but from an adoption view, i think there's a lot of
> work to be done.

There always is.

Thanks for the thoughts,

- C


-- 
You received this message because you are subscribed to the Google Groups 
"pylons-discuss" group.
To post to this group, send email to [email protected].
To unsubscribe from this group, send email to 
[email protected].
For more options, visit this group at 
http://groups.google.com/group/pylons-discuss?hl=en.

Reply via email to