ORM? DAL?

274 views
Skip to first unread message

Massimo Di Pierro

unread,
Jul 12, 2008, 5:47:58 PM7/12/08
to web...@googlegroups.com
Jonathan Ellis from SQLAlchemy has suggested the we do not call the
web2py ORM and ORM since it does not map Python classes into SQL
tables. While I still think that web2py does fit the wikipedia
definition of ORM, I do not particularly care about the name and I
actually think that calling "it" Database Abstraction Layer (DAL)
instead of ORM may actually make us more marketable.

What do you think?

Massimo

ockhamwasright

unread,
Jul 12, 2008, 6:47:54 PM7/12/08
to web2py Web Framework
"O, be some other name! What's in a name? that which we call a rose by
any other name would smell as sweet;" ~ William Shakespeare, Romeo and
Juliet, 1594

You will hopefully pardon the melodrama, but I think the people who
will benefit most from Web2Py and will drive its wide-scale acceptance
wouldn't care whether you called it "ORM", "DAL" or "Albanian lesbian
sushi chef". If the techno-weenies care about such matters, then fine.

yarko

unread,
Jul 12, 2008, 7:32:09 PM7/12/08
to web2py Web Framework
I think Jon Ellis is right - and frankly, WIkipedia, while stating
this abstractly, I think says the same...

Check out http://en.wikipedia.org/wiki/Database_abstraction_layer and
ask "which better describes what we're doing in web2py?"

I think what we have in Web2Py is, indeed, a DAL.

I haven't (recently?) read Scott Ambler's discussion in
http://www.agiledata.org/essays/mappingObjects.html ...

If anything, web2py _not_ providing something as structured as an ORM
(among other things) avoids the performance problems this can
typically introduce.

What do you think?

yarko

unread,
Jul 12, 2008, 7:34:09 PM7/12/08
to web2py Web Framework
Wikipedia ORM link, so you can quickly compare with DAL:
http://en.wikipedia.org/wiki/Object-relational_mapping

On Jul 12, 6:32 pm, yarko <yark...@gmail.com> wrote:
> I think Jon Ellis is right - and frankly, WIkipedia, while stating
> this abstractly, I think says the same...
>
> Check outhttp://en.wikipedia.org/wiki/Database_abstraction_layer and

Stuart Rackham

unread,
Jul 13, 2008, 12:57:09 AM7/13/08
to web...@googlegroups.com
+1

Massimo Di Pierro

unread,
Jul 13, 2008, 11:55:01 PM7/13/08
to web...@googlegroups.com
OK, let DAL be!

I will fix it.

Massimo

Joe Barnhart

unread,
Jul 14, 2008, 1:03:39 AM7/14/08
to web2py Web Framework
I updated the Wikipedia page. It still needs filling out a bit.
Maybe I'll get to it this week since the page is "sticking" this time
and hasn't done its disappearing act!

-- Joe B.

Massimo Di Pierro

unread,
Jul 14, 2008, 1:15:00 AM7/14/08
to web...@googlegroups.com
Thank you Joe.

Massimo

yarko

unread,
Jul 14, 2008, 1:22:21 AM7/14/08
to web2py Web Framework
I noticed it's sticking!

blackthorne

unread,
Jul 14, 2008, 1:34:37 AM7/14/08
to web2py Web Framework
This is getting bizarre.
I was producing a document to send you with all I think about web2py.
I was writing exactly that at this precise moment, and actually did it
in other thread.
Yes, not that I care about the name but web2py brings a DAL not so
much a ORM, i agree with that.

I think the naming strategy depends on you...
Consider this: With a DAL it's easy to take it to the next step, an
ORM. It's something really easy to achieve (with no backward
compatibility break).

If you want to make/delegate this next step keep calling it ORM and
consider this just a stage of the process until getting there. Else,
use DAL.

Now the final conclusion seems twisted to me... most people that just
see web2py for the first time, as soon as they find that there is no
ORM probably they won't even bother guessing what a DAL is, and what
can they do with that. People love ORM's, that's on the high, it's not
something you want to run away from...

Francisco


On Jul 12, 10:47 pm, Massimo Di Pierro <mdipie...@cs.depaul.edu>
wrote:

Kyle Smith

unread,
Jul 14, 2008, 2:26:43 AM7/14/08
to web...@googlegroups.com
Data Abstraction Layers have been around for ages. While it might be deceiving for new users who are comparing web2py to other currently trendy frameworks to see DAL instead of ORM, developers that actually understand the two concepts won't have a problem. Perhaps a small entry on the page about the DAL and the ways in which it is different from an ORM would be helpful. There's already an AlterEgo entry about the differences between web2py's DAL and SqlAlchemy.

Kyle
Message has been deleted

yarko

unread,
Jul 14, 2008, 2:38:38 AM7/14/08
to web2py Web Framework
"A rose by any other name would [still be a rose]...."

corollary:

'tis best to call a rose a rose...


On Jul 14, 1:26 am, "Kyle Smith" <kyletsm...@gmail.com> wrote:
> Data Abstraction Layers have been around for ages. While it might be
> deceiving for new users who are comparing web2py to other currently trendy
> frameworks to see DAL instead of ORM, developers that actually understand
> the two concepts won't have a problem. Perhaps a small entry on the page
> about the DAL and the ways in which it is different from an ORM would be
> helpful. There's already an AlterEgo entry about the differences between
> web2py's DAL and SqlAlchemy.
>
> Kyle
>
> On Sun, Jul 13, 2008 at 10:34 PM, blackthorne <Francisco....@gmail.com>

blackthorne

unread,
Jul 14, 2008, 3:01:08 AM7/14/08
to web2py Web Framework
I will put it this way. What I am about to say has ocurred, it's not a
theorical demonstration based on theorical hypothesis.

I was on IRC, trying to learn more about Django and talking about it's
web interface functionality.I was looking for something close to
web2py in the 3rd party world of Django.
Someone said it was impossible, and I used web2py as an example to
proof it was possible.
There was a fellow, that actually searched for the web2py official
website and closed it as soon as he found the slogan "web enterprise
framework".
His problem is on the "enterprise" word.
( http://oebfare.com/logger/django/2008/07/12/ - 23h:15m )

Now you can think he is @#)( and a !@$R@#$ and I agree with you but
this is NOT an irrelevant issue.
Thing is, open source community (our target) is closely tied with
communism. Just thinking about Open Source concept and Communism and
it's easy to understand why... now, I'm not trying to open a political
debate on this and I'm not saying that a member of Open Source
community has to be a communist. I'm just saying that there is a
connection between those two often noticeable in things like this:
"hate for enterprises" and "close source is always worse",...

Now consider removing the "word" Enterprise from the slogan and a
closed tab becomes a rose!

Seriously, this marketing thing makes magic. Being better sometimes is
the less relevant thing to achieve sucess.


On Jul 14, 7:31 am, yarko <yark...@gmail.com> wrote:
> dinner & beer satisfied (but no hot bath - caveat emptor!)
> ...
>
> consider these separately....
> --- wants (ORM?)
> --- fears  (no ORM, no people)
> --- thoughts (i.e. expectations - "if you have ORM it is better")
> --- data (what is)
>
> if "People love ORM's...." and "as soon as they find there is no
> ORM...." - is that as true if told there is an ORM, and discover it's
> not?
> (Or will people be sold, hollywood style, and not really know what it
> is - so long as they are told "...isn't the king's new ORM
> _beautiful_!" ?  I think people are smart).
>
> A rose by any other name is still a rose ...
>
> corollary:
> best to call a rose a rose...
>

Joe Barnhart

unread,
Jul 14, 2008, 4:28:45 AM7/14/08
to web2py Web Framework
Regarding "enterprise"...

This is kind of "marketing 101". A descriptive term like "enterprise"
is often thought of as enhancing the image of a product, but it can
also _limit_ its market. Suppose someone is looking for a web
framework for small business or a small non-profit? They might skip
something called "enterprise" thinking it is only for very large
sites.

-- Joe B.

P.S. What are the best traits of web2py? That's where you start when
looking for the "value proposition" of the product. Some
observations:

-- It's the most pythonic framework I've seen. It actually IS
"turtles" (python) all the way down.

-- It is compact, fast, and "batteries included" (server, dal,
templating, etc., etc...)

-- It makes web programming obvious (which is a _good_ thing)

-- Simple websites are simple, complex websites are only a little
harder

yarko

unread,
Jul 14, 2008, 9:36:57 AM7/14/08
to web2py Web Framework
It sounds a little like you are coming to the Rose and saying "ahhh,
if you were as fragrant as the lilac, then you would be a grand
flower..."
...and going to the lilac, and saying "why are you not more like a
Rose...."

and both blooms are closing in response...

There is something about shame - comparing to point out what is "less"
or who is "better"....

Compare, instead, to learn..

Every project has something to offer, and something to learn from...
and every project makes choices that will not be the same as choices
of another project.

Consider: "I really like how xxx solved this. Here's what I've
learned from that, and I'd like to try to apply what I've learned
here...."

(The "Zen" of open source? ;-)

On Jul 14, 2:01 am, blackthorne <Francisco....@gmail.com> wrote:
> I will put it this way. What I am about to say has ocurred, it's not a
> theorical demonstration based on theorical hypothesis.
>
> I was on IRC, trying to learn more about Django and talking about it's
> web interface functionality.I was looking for something close to
> web2py in the 3rd party world of Django.
> Someone said it was impossible, and I used web2py as an example to
> proof it was possible.
> There was a fellow, that actually searched for the web2py official
> website and closed it as soon as he found the slogan "web enterprise
> framework".
> His problem is on the "enterprise" word.
> (http://oebfare.com/logger/django/2008/07/12/- 23h:15m )

Massimo Di Pierro

unread,
Jul 15, 2008, 12:00:40 AM7/15/08
to web...@googlegroups.com
love this!

On Jul 14, 2008, at 1:31 AM, yarko wrote:

>
> dinner & beer satisfied (but no hot bath - caveat emptor!)
> ...
>
> consider these separately....
> --- wants (ORM?)
> --- fears (no ORM, no people)
> --- thoughts (i.e. expectations - "if you have ORM it is better")
> --- data (what is)
>
> if "People love ORM's...." and "as soon as they find there is no
> ORM...." - is that as true if told there is an ORM, and discover it's
> not?
> (Or will people be sold, hollywood style, and not really know what it
> is - so long as they are told "...isn't the king's new ORM
> _beautiful_!" ? I think people are smart).
>
>
> A rose by any other name is still a rose ...
>
> corollary:
> best to call a rose a rose...
>
>
> On Jul 14, 12:34 am, blackthorne <Francisco....@gmail.com> wrote:

Reply all
Reply to author
Forward
0 new messages