interpreter, just that BD has an advantage over MM CF in this particular
case because it uses an "interpretive approach" as you like to call it.
I'm not quite sure how "interpretive approach" differs from being an
interpreter.. the explanation I received is that BD produces a tree
structure of all tags in memory and on each call iterates this tree
structure to call the classes and methods corresponding with each tag.
Sounds like an interpreter to me, but I don't actually have a CS degree so
perhaps I'm missing some distinction.
Also, being compiled doesn't necessarily mean faster either, 'cause most
processing takes place in the code assiociated with the tags itself (written
by MM and New Atlanta) as opposed to the code generated/interpreted by the
CFML developer. From what I saw the BD interpreter was extremely fast,
although I've never run any empirical speed comparisons.
Of course, I don't see any way BD will ever have the Blackstone features
already announced..notably the Flex integration (although I personally don't
think this is a valuable feature), so we are likely to see framentation in
the future and products can compete based on features and people will
develop code that only runs on one "flavor" of CF, which is what I always
saw as the problem with BD--it's always playing catch-up and any innovation
it does just causes fragmentation.
My $0.02.
Sam
-----Original Message-----
From: Matt Liotta [mailto:[EMAIL PROTECTED]
Sent: Thursday, June 10, 2004 3:24 PM
To: CF-Talk
Subject: comparing CF and BD (was Re: COLDFUSION has JUST been
DISContinued!!!!)
> This gives New Atlanta a huge advantage over Macromedia in that since
> MM's
> CF engine produces Java byte code directly it'd be nearly impossible
> for MM
> to port to .NET without re-writing. Remember that New Atlanta's
> engine is
> an interpreter and not a compiler.
>
I would like to make a clarification that I think is important when
comparing CF and BD. The CF engine does use a compiler approach that
produces Java byte code. BD however isn't strictly an interpreter. It
certainly uses an interpretive approach, but like CF, all of the tags
are Java classes created ahead of time. Thus, you could think of BD as
an interpreter for basic logic, but the tags's logic is actually
compiled.
Anyway, the point I want to make is that comparing the two products
from an approach point of view is more complicated than it first
appears.
-Matt
[Todays Threads] [This Message] [Subscription] [Fast Unsubscribe] [User Settings]
