tracker issue : CF-3038965

select a category, or use search below
(searches all categories and all time range)
Title:

Bug 78067:(Watson Migration Closure)Bring back CFC-Beans

| View in Tracker

Status/Resolution/Reason: Closed/Won't Fix/

Reporter/Name(from Bugbase): Devin Holloway / Devin Holloway (Devin Holloway)

Created: 06/21/2009

Components: Language, CF Component

Versions: 9.0

Failure Type: Unspecified

Found In Build/Fixed In Build: 0000 /

Priority/Frequency: Normal / Unknown

Locale/System: English / Platforms All

Vote Count: 9

Problem:

Bring back CFC-Beans. These could be the single most performance enhancing features of CF9 and help to put the notion of, "CF's performance is too slow to use it in this situation", to rest. This feature would also be well received by mid-higher level CF developers as well as Flex/AS3 developers making a decision on a suitable backend for their Flash applications.

A possible altnernative is to use a completely new tag, such a <CFBean>, <CFValueObject>, or any other possible implementation that results in the same performance gains.

This enhancement is listed as a severity of 2 (Seriously impedes use of product) because it is a deal breaker in many performance-related situations that require unfavorable work-arounds or dropping CF entirely for another language.
Method:


Result:

----------------------------- Additional Watson Details -----------------------------

Watson Bug ID:	3038965

External Customer Info:
External Company:  
External Customer Name: Devin Holloway
External Customer Email: 5E060FD6422E831A992015D5
External Test Config: 06/21/2009

Attachments:

Comments:

I agree, +1.fffffffvcf I think the requiremnt for "fast" CFC's is definitely there. I have never had a problem with the "strict" attribute. I think it could stay the way of its original implementation. There are plenty of examples whereby an attribute changes the behaviour completely of a construct in any programming language. But if it helps get this back into the product then a name change to CFCBEAN is fine with me too. As for the severity... I think 2 is a good value to assign it. Along with Adam, I think CFML is mature enough now for its users to be allowed to expect "faster" CFC creation. At the very worst, you give it a new name so as to avoid any confusion with "normal" CFCs and it doesn't get used a lot. Since you already have it developed - is that really an issue? The upshot of course, is that those who want to make use of it can. Gavin.
Vote by External U.
23606 | November 10, 2011 06:54:23 PM GMT
+1 I don't agree that it should have been scuttled. As Adam C said, waiting for another product release is a very long time to wait for performance improvements. A fast value object (with an IBO) is a very doable performance improvement over traditional getters and setters. We need this now. It might be too late two years from now.
Vote by External U.
23607 | November 10, 2011 06:54:24 PM GMT
+1 vote. Not sure about the <cfvalueobject> suggestion, as it's a bit... err... "verbose", maybe. <cfbean> is fine, and to the point. I even think <cfcomponent strict="true"> is fine, to be honest. Agree with the severity of the issue, because if the Adobe bods themselves initially implemented the "strict" notion to get over a performance hump, then there's clearly a performance hump there needing to be surmounted. Down the track I suppose we're going to want to solve this perceived performance issue with *all* CFC instantiation, but this is a good v1.0 step (or, let's be honest... they've had multiple goes at it already, with 6.0, 6.1, 7.0, 8.0, 9.0... this'd be v5.0 as far as CFCs are concerned). I'd prefer this interim - but not ideal - solution to be implemented now as an intermediary step, than have to wait around yet another version to get the complete/comprehensive solution nailed. I do not accept that there will be this perceived community confusion that has been touted as the reason for dropping this from v9.0. Whether I'm right or wrong, clearly the topic needs further discussion than just an arbitrary SP5-ing of the concept. At least I'd like to hear the "nae" argument from the bods that think there'll be this confusion. So we can discuss it. -- Adam
Vote by External U.
23608 | November 10, 2011 06:54:25 PM GMT
+1 vote, also +1 vote for a new CFBEAN type. This is a *must* - we've been waiting a long time to get fast CFCs for things like Flex Remoting so please do not drop this feature!
Vote by External U.
23609 | November 10, 2011 06:54:26 PM GMT
I was really disappointed when Rupesh reported that it was been dropped. In beta 2 release, this was the greatest enhacement added to the product. Finally we could get very high speed beans to send back and forward between CF and Flex. Very often we listen Adobe reporting CF as the perfect backend solution for Flex which is totally false, till we add this feature. Other SS solutions are much faster than CF is, unless you use typed structures, which is just an hack, nothing more.
Vote by External U.
23610 | November 10, 2011 06:54:27 PM GMT
This bug has been voted..
Vote by External U.
23611 | November 10, 2011 06:54:28 PM GMT
We (Flex Developers) need CFC bean!!
Vote by External U.
23612 | November 10, 2011 06:54:29 PM GMT
This bug has been voted..
Vote by External U.
23613 | November 10, 2011 06:54:30 PM GMT
This bug has been voted..
Vote by External U.
23614 | November 10, 2011 06:54:31 PM GMT