Patrick, I think it would be a great idea if you were to change the core file to do what you suggest and then run some performance tests. I personally don't use many of the Fusebox variables (although I would want to think and discuss it more before I advocated removing them). Still, if significant performance increases were possible, this would surely buttress your argument.
-----Original Message----- From: Patrick McElhaney [mailto:[EMAIL PROTECTED]] Sent: Friday, April 12, 2002 1:54 PM To: [EMAIL PROTECTED] Cc: [EMAIL PROTECTED] Subject: Circuit aliases discussion >>> from the fbx_settings thread on [EMAIL PROTECTED] Adam, I've been trying to get homeCircuit, targetCircuit, and thisCircuit ripped out of the spec ever since they were put in. IMHO, a circuit should not make any decisions based on the value of its alias, and certainly not the values of any other aliases in the application, because those names belong to the application and the circuit has no control over them. If one of these variables is compared to a hard-coded string, the circuit will break as soon as the alias assigned to that circuit changes. The only reasonable uses for these variables are as follows: 1) To compare them to each other. For example, we could use (fusebox.thisCircuit eq fusebox.homeCircuit) to determine whether the current circuit is also the home circuit. 2) To ensure the same circuit is called in the next request without requiring that the circuit know the value of its alias. <cfset xfa.doThis = "#fusebox.thisCircuit#.doThat> #1 was abstracted into two booleans, fusebox.isHomeCircuit and fusebox.isTargetCircuit. #2 can be accomplished with fusebox.circuit instead. Now let's look at the reasons to remove the variables: 1) It would become impossible to write code compares a circuit's alias to a hard-coded string. 2) We wouldn't need to do the reverse-lookup, which is probably the most expensive operation in the core file. 3) It would be easier to implement the specification. 4) Fewer things for a new Fuseboxer to learn and understand. A couple of days ago, Hal mentioned that Fusebox should be target to the center of the bell curve of user sophistication, as described in Alan Cooper's book, The Inmates are Running the Asylum. (I think everyone on the standards committee should read that book, BTW.) Cooper does that mostly by removing features. It's a recurring theme that comes up in just about every page of the book. John Q and I have argued about this topic countless times and I don't think we're any closer to agreement than Sharon and Arafat. I'm hoping that this time some of the others on the list will really give it some thought and weigh in on the subject. Patrick ==^================================================================ This email was sent to: [email protected] EASY UNSUBSCRIBE click here: http://topica.com/u/?bUrFMa.bV0Kx9 Or send an email to: [EMAIL PROTECTED] T O P I C A -- Register now to manage your mail! http://www.topica.com/partner/tag02/register ==^================================================================
