> Hi Patrick,
>
> My main reason for using circuits.cfm is to shield the developer
> from stuff
> that he does not need or want to know. If I want to call the "login"
> sub-app, I just need to know that its name is "login". I simply
> DO NOT want
> to know that it is actually located at /apps/library/security/login. Even
> worse, I don't want to have to change all of my code when the system
> architect decides to move /apps/library/security/login to
> /apps/library/cfml/security/login.
I don't see a reason to go moving circuits around after you've snapped
them together. Suppose I have a circuit that looks like this:
cart/
cart/shop/
cart/checkout/
cart/orderTracking/
cart/admin/
Basically, it's a standard shopping cart app, right? It doesn't make
sense to put all of its many functions in one circuit, so I broke
out the four basic use cases into nested circuits. This app
functions as a stand-alone, but the real intention is to nest it in
a larger application.
Once the "cart" application is nested in a larger application do
its five circuits become independent of each other? Would
it make sense for me to move the "checkout" circuit to another
part of the larger application? Would it still be valid outside
of "cart"?
Obviously, it's very important to you to be able to move circuits
around after they've been snapped together -- and furthermore that
this can be done in a routine fashion -- but I just don't see
the motivation for it. Having mulled over it for days, I still
can figure out what I'm missing.
> in short, there ARE great
> advantages to
> using a two-part name.
Advantages? With an 's'? So what are the others? :)
> Circuits.cfm has other side-benefits, eg it provides and centralizes a map
> of the system.
That's a circular argument. The only map it provides is the one
it creates a need for. :)
> So let us throw you the obvious reply:
>
> Why would you WANT to refer to a circuit by its fully-qualified name?
1) It's a simpler, more direct approach to the problem. Its kind of like
saying, "I'm from USA.NC.CHARLOTTE" vs. "I'm from City Code 'CLT' and here's
a spreadsheet that maps city codes to their nation.state/province.fullname."
That's the main reason. When all other things are equal, the KISS
principle applies.
2) I don't want to bother with creating and maintaining another file
(circuits.cfm) if I don't have to.
3) Context matters. Store.Cart.Admin makes more sense to me than
'admin', 'adm', or whatever. (And no, you can't call it
'store_cart_admin', because that defeats the premise of abstracting
away the location.)
4) I can literally drag and drop new circuits into an app and have
them work without any changes.
5) I don't have to worry about confusing two different circuits with
similar names.
6) Makes it impossible to violate a circuit's encapsulation. A circuit
should only be able to directly refer to itself and its descendants.
(Of course, it can have XFA's passed up from its ancestors that refer
to circuits outside of its scope.)
7) If I introduce a new circuit with the same name as an existing
circuit, I'm don't have to rename it. And as a result, I don't have
to go and find any fuseactions that point to the old name and update
the code.
8) A special variable would refer to the current circuit.
(for example, <cfset xfa.submitForm = "#this#.addProduct">) That
prevents me from having to remember what the current circuit's
name is, and makes it clear when a circuit is referring to itself.
Okay guys, the ball's in your court. What are the other advantages
to two-part names? Let's weight the pros and cons.
Patrick
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Structure your ColdFusion code with Fusebox. Get the official book at
http://www.fusionauthority.com/bkinfo.cfm
Archives: http://www.mail-archive.com/[email protected]/
Unsubscribe: http://www.houseoffusion.com/index.cfm?sidebar=lists
