***************************** Team Allaire *****************************
Sure. Just as a side on this, my standards of testing is the following:
<CFSET start=gettickcount()>
<CFLOOP from="1" to="#loopcount#" index=i>
code here
</CFLOOP>
<CFSET end1=gettickcount()-start>
The first run is never used as this is the 'put into P-code' run. After that
I simply watch the times as they go by. A more advanced version basically
sets the info into application variables and gets an average of performance
over time.
> Woo-hoo! Excellent. Thanks, Russ. Michael D., can you oversee the metrics
> testing to make sure our methodology isn't flawed?
>
> -----Original Message-----
> From: Johnson, Russell [mailto:[EMAIL PROTECTED]]
> Sent: Tuesday, May 22, 2001 12:20 PM
> To: Fusebox
> Subject: RE: Why Benchmark? (An answer to Patrick)
>
>
> I have taken the Content management system example application that comes
> with CF Studio and converted it to FuseBox. I didn't rewrite any of the
> internal code, only separated it out into fuses and put it into XFB. You
> guys are more than welcome to use it if you want. That way you have the
> non-fusebox written by Allaire and a Fuseboxed version.
>
> Does it matter if we test regular FuseBox or XFB or both?
>
> Russ Johnson
> Web Environment Services
> Blue Cross Blue Shield of Florida
> Work: 904-905-7130
> Mobile: 904-534-8162
> [EMAIL PROTECTED]
>
> "Great thinkers have always encountered violent opposition from mediocre
> minds." -- Albert Einstein
>
>
>
>
> -----Original Message-----
> From: Hal Helms [mailto:[EMAIL PROTECTED]]
> Sent: Tuesday, May 22, 2001 11:38 AM
> To: Fusebox
> Subject: RE: Why Benchmark? (An answer to Patrick)
>
>
> ***************************** Team Allaire *****************************
> I've been thinking about this and it seems to me we should take an
> application that's already done in non-Fusebox and then rewrite it with
> Fusebox. By doing this, we can avoid the argument that the non-Fusebox app
> was a straw man, written in order to fail. Does anyone have access to any
> non-trivial applications that are well-written and are not already
> Fuseboxed?
>
> -----Original Message-----
> From: Judith Dinowitz [mailto:[EMAIL PROTECTED]]
> Sent: Tuesday, May 22, 2001 11:16 AM
> To: Fusebox
> Subject: Why Benchmark? (An answer to Patrick)
>
>
>
> > > I agree with you on most points. I'm not converting to Fusebox for
speed
> > > increase. More portability and make my site more manageable so I can
> > > concentrate on adding cool features.
> > Exactly. So why are you benchmarking?
>
> Look, we all agree that FB is a great methodology, or we wouldn't be here
> and we wouldn't be supporting it.
> So, you ask a good question: Why benchmark?
>
> Wouldn't you like to have some scientific proof, some statistics to give
to
> people who nay-say FB? Also, wouldn't you like to be able to pinpoint
which
> parts of the FB methodology are faster and which may add a little time?
> Maybe this way we can make the whole thing more efficient. We can continue
> to refine the methodology.
>
> The point of this is not to say "Plain CF is better" or "FB is better."
The
> point is to prove that one does not _lose_ significantly in processing
speed
> when converting to Fusebox. And if there are speed bottlenecks in the
> process, let's pinpoint them and come up with solutions.
>
> How many times have you heard someone say "FB has too much overhead!"
Well,
> let's prove them wrong. But let's make sure these tests are done well, and
> look at this whole thing scientifically. I think the whole community will
> benefit.
>
> Judith
>
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
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
