djWarehouse - Django based e-commerce system

[alex.v...@gmail.com]

Hello,

I am pleased to announce the immediate availability of djWarehouse e-
commerce system, developed on Django.

It is called djWarehouse and there are Trac site for it:
http://www.djwarehouse.org/

You can see demo site (with open /admin/ area) here:
http://demo.djwarehouse.org/

Here is instructions how to install "simple demo site" under Linux/
FreeBSD:
http://djwarehouse.org/wiki/DemoInstall

Regards,
Alex

[Nicolas Steinmetz]

alex.v...@gmail.com a écrit :

> Hello,
>
> I am pleased to announce the immediate availability of djWarehouse e-
> commerce system, developed on Django.

Could you provide a quick bench / swot against satchmo project as it
looks you have the same objectives (being an e-commerce platform).

Regards,
Nicolas

[alex.v...@gmail.com]

On Feb 18, 4:10 pm, Nicolas Steinmetz <nsteinm...@gmail.com> wrote:
> > I am pleased to announce the immediate availability of djWarehouse e-
> > commerce system, developed on Django.
>
> Could you provide a quick bench / swot against satchmo project as it
> looks you have the same objectives (being an e-commerce platform).

I've tried to explain why we did not used Satchmo in our first site
we've build:
http://djwarehouse.org/wiki/About (read at bottom).

What about current differences ? I am not 100% sure because its
difficult to quickly review
Satchmo code, but here is what I've noticed at first look:

djWarehouse itself is not a e-commerce shop, instead its a tool to
create different e-commerce stores. To make an
example, we took djWarehouse and created simple demo store (http://
demo.djwarehouse.org) but that is very minimalistic example. We plan
to release a 'bike shop' and 'gadget shop' examples later, which will
demonstrate
the power of using 'product attribute groups' and 'calculated
attributes'.

I think the most differences are:

1. We try to keep djWarehouse very minimalistic. Satchmo, by design
have everyting in it. This is true about core model fields and total
amount of core applications installed by default. We think that
developer would decide what is required, and add specific modules on
demand.

2. We tried to make system as pluggable as possible. For example you
can have own Shipment,Payment,Discount,Promotion modules which are
defined in absolutely different manners. Product Catalog, Customer and
other things can be easily extended by specific fields (via OneToOne
models of Django).

We don't hardcode many things in core, except a very simple, common
ones (as for example RegionalTax which is common in US). We don't
assume a things, like 'credit card payment' would be needed, so its
not added by default.
We think that someone, who makes store configuration, would add a
required modules.

When you look at admin site, you can compare it to Satchmo.

Please also note, I think one of important differences is that Chris
Moffit is much better communicator then we are, I expect he is
natively speaks English, and he makes a good advertisement. Also it
seems that he spends some
amount of time into making Satchmo popular. We don't have enough time
for that, since we are small company
and all we try to do is make more e-commerce stores using djWarehouse,
and document how we do it.

We were afraid to show something which is half ready. Now it looks
like we are almost ready, so we try to build sites on djWarehouse, and
see how it works, and document things which worked well on our site.

Regards,
Alex

[Chris Moffitt]

I'll try to respond to some of these points. To be fair, I haven't spent much time looking at the internals of djWarehouse. From the surface, it seems like they have some good experience and design going into the product. I'll point out some differences - but they are mostly based on looking at the site and not actually playing with the code:

- Satchmo is a community app while djWarehouse is built by a company (but released open source) and used for their needs. There are at least 5 Satchmo stores in production and probably more out there that I don't know about. Our mailing list has over 400 users and the dev list has about 75 people. We do get lots of patches and bug fix suggestions from the community but there are only 4 of us with commit privileges. djWarehouse has more of a dedicated team. I don't know that one model is better than the other - they are different.
Satchmo has addition features such as:
-  payment modules for Authorize.net, Paypal, Google, cybersource and trustcommerce. It also has a COD module and auto success module which might be useful for certain types of sites.
-  UPS integration for estimating shipping charges.
- support for multiple product types including gift certificates, downloadable products, custom products, etc.
- Satchmo allows you to generate PDF invoices and packing slips.
- supports multiple shipping calculation methods and includes geographic info for most countries.
- Satchmo allows translation of content and allows you to test the different translations.
- Satchmo has extensive unit tests
- There are probably others but these are some of the things that come to mind immediately.

To be fair, Satchmo has 2 main warts:
- The product models are not as extensible as we'd like. It can be done but it could be cleaner. I'm hopeful that the new model inheritance code will help us out.
- The admin interface could be slicker. Our plan is to tackle that as part of the newforms admin merge.

There's also a constant struggle with building "frameworks." Trying to decide what does and does not go in is hard. Sometimes I think we hit it right and other times I'm not sure. I think the djWarehouse crew felt we had too much in the framework. It's a fair comment and reasonable people may disagree on this point.

I will also freely admit that I am a native english speaker (and do not speak any other languages with any proficiency whatsoever) and that I do try to spend time promoting Satchmo ;) My main goal is to let people know there is a reasonable alternative to OsCommerce! I don't want to put down djWarehouse or other projects. I suspect there is a lot we can learn from their project and would welcome any specific criticisms or suggestions for improving Satchmo. It's a big internet out there and there are a lot of different needs. I'd prefer to have 1 framework out there to consolidate efforts but sometimes that just isn't possible.

If someone has played with both versions and has any more specific details, feel free to let me know.

-Chris

[alex.v...@gmail.com]

Thanks Chris for joining this discussion ;-)

On Feb 18, 10:49 pm, "Chris Moffitt" <ch...@moffitts.net> wrote:
> - Satchmo is a community app while djWarehouse is built by a company (but
> released open source) and used for their needs. There are at least 5 Satchmo
> stores in production and probably more out there that I don't know about.

Chris, you have done an excellent work to promote
Satchmo. In fact all you say is very right. I think that getting
application to community is not easy thing - it probably takes as
twice time as to produce source code.

> Our mailing list has over 400 users and the dev list has about 75 people. We
> do get lots of patches and bug fix suggestions from the community but there
> are only 4 of us with commit privileges. djWarehouse has more of a dedicated
> team. I don't know that one model is better than the other - they are
> different.

Yes. In fact, when we started developing djWarehouse, we known
*exactly* what we need
because we already have some customers, and our main goal was to
satisfy their needs.

We've finished with a first task, now we are releasing our code to
community to find
out if that code is useful to them, as well.

> Satchmo has addition features such as:
> - payment modules for Authorize.net, Paypal, Google, cybersource and
> trustcommerce.

We do have Google Checkout as free module (additionally we can sell
Protx module as commercial one).

> It also has a COD module and auto success module which might
> be useful for certain types of sites.
> - UPS integration for estimating shipping charges.

djw: UPS

> - support for multiple product types including gift certificates,
> downloadable products, custom products, etc.

djWarehouse is different design. We do have very basic product, and
then you can customize
it the way so it can be downloadable, or any other kind.

> - Satchmo allows you to generate PDF invoices and packing slips.

We have developed pdf generation framework based on iReport and
reportlab. This framework
allows to draw any kind of graphical reports in iReport and produce
invoices & etc. While we did not
considered it as a part of djW (since not everyone needs it). On demo
store invoices are plain HTML,
and printed from there. PDF generation has been used in one of sales
processing databases we've
created for one e-commerce customers. It is available as a part of
DjHDGutils here:
http://public.halogen-dg.com/wiki/DjHDGtools

> - supports multiple shipping calculation methods and includes geographic
> info for most countries.

Yes, Django regional data is very good. We don't have such
in djW, but I think its a pluggable thing, probably someone can use
Satchmo part
for this part of code, if required.

> - Satchmo allows translation of content and allows you to test the different
> translations.

Yah, that is the thing which we decided not to go in by default
because, according
to our experience only around some people need that, while it could:
1) make admin screens wider due to additional fields & controls
2) depending on design - make software slower (due to additional
language queries).

> - Satchmo has extensive unit tests

Agree, we don't have many. I've tried to develop some, but decided
that it is impossible to test
something which is half-working. DjWarehouse itself is not a store,
its a framework
to build stores. As result, it does not work by itself - only after
customizing. While djW itself does not have
a lot of tests, sites based on djW do have unit tests.

> - There are probably others but these are some of the things that come to
> mind immediately.

Thank you for your report ;-)

> To be fair, Satchmo has 2 main warts:
> - The product models are not as extensible as we'd like. It can be done but
> it could be cleaner. I'm hopeful that the new model inheritance code will
> help us out.

Yah, I also hope that inheritance would help. Until then, we are
customizing data via OneToOne relational
tables, which also works greatly.

> - The admin interface could be slicker. Our plan is to tackle that as part
> of the newforms admin merge.

Is newforms-admin is anywhere ready? Last time we've tried it, it
missed a lot
of things. At all, we are developing our own admin suite which
probably would
work much better then newforms-admin.

> There's also a constant struggle with building "frameworks." Trying to
> decide what does and does not go in is hard.

In fact I belive that its impossible. Since we've built many e-
commerce stores, I would say
that neither of solutions I've seen, including our own will satisfy
all needs.
We are trying to come up with the reasonable default, according to our
own
experience, but I know the cases when neither of those ready patterns
will work.

> Sometimes I think we hit it
> right and other times I'm not sure. I think the djWarehouse crew felt we had
> too much in the framework. It's a fair comment and reasonable people may
> disagree on this point.

Agree. That is a matter of a e-commerce store needs. If Satchmo
matches typical
user needs better, then people would think its easier to use it. While
nothing
prevents us to release Satchmo-like configuration for djW ;-)

> I will also freely admit that I am a native english speaker (and do not
> speak any other languages with any proficiency whatsoever) and that I do try
> to spend time promoting Satchmo ;) My main goal is to let people know there
> is a reasonable alternative to OsCommerce!

Yes and you do it very good. ;-) The more people comes to Django, the
better!

> I don't want to put down
> djWarehouse or other projects. I suspect there is a lot we can learn from
> their project and would welcome any specific criticisms or suggestions for
> improving Satchmo.

I still think we need to discuss and try to implement an independent
plugin
architecture for the 3rd party modules - Payment, Shipment. I have a
half
ready USPS, we already have UPS (you too) and I think we are spending
our time
wrong way, by trying to develop same modules twice. Its my problem
that I
suggested it and did not come with a reasonable offer. I will do it in
a short time - will
look at Satchmo code and will try to publish a document, explaining
how to
make pluggable modules to work. That is to benefit both of our
solutions, and mostly to benifit Django as framework.

> It's a big internet out there and there are a lot of
> different needs. I'd prefer to have 1 framework out there to consolidate
> efforts but sometimes that just isn't possible.

Agree. There are no ideal soluition for everyone.

> If someone has played with both versions and has any more specific details,
> feel free to let me know.

Also would be interesting to hear.

Thanks, Chris!

[Nicolas Steinmetz]

Thanks for your reports. I'll keep an eye on both then :

It's nice to see such a positive attitude from their lead
developper/promoter when comparing similar products :-)

Nicolas