Dave, I hope you will excuse me jumping on but you give some great insights here and as this thread is old I thought IT might be ignored so I did this, because this makes good reading for all of us.
Kind Regards - Mike Brunt, CTO Webapper http://www.webapper.com Downey CA Office 562.243.6255 AIM - webappermb "Webapper - Making the NET work" -----Original Message----- From: Dave Watts [mailto:[EMAIL PROTECTED]] Sent: Thursday, September 19, 2002 9:18 AM To: CF-Talk Subject: RE: Jsp Vs Cfm (CFMX) -- Test Code > > always strongly recommended the use of a compiled language > > for the "middle tier" - you'd only do presentation logic > > in ASP, and you'd do anything of any complexity in COM > > with either C++ or VB. I suspect that's a little different > > with .NET, but I'm not sure how different it is. > > FYI... .NET/C#/VB.net gets compiled into MSIL and the runtime > runs the code "compiled* not interpreted. Com's in .NET are > supposedly slower. Classes are the new standard of development. My suspicion is about how to segment code in .NET, not about the compilation process. In "classic" ASP, you'd do presentation logic in ASP, and do heavy lifting in COM. I'm not sure what the MS best practices are with ASP.NET - do you write all your logic in aspx pages, or do you still write separate components? If you write separate components, does this give you any performance benefit, such as the components being managed in whatever they call MTS/COM+ these days? > > ... considering that CF has to work with the huge volume > > of legacy code out there. > > Microsoft VB code does the same thing (Optional Type > declarations) as serveral others pointed out here. Uh, yes, but that's the way the language was originally constructed. CF wasn't constructed that way. If you've ever worked with VB, you'll know that MS has occasionally broken compatibility between versions, also. > If it takes considerably more time for a million loops.... > that means it will take more time to run ONE LOOP than the > counter Web Application Server/Language. We are doing a > million loops here as a CASE/Example...to prove the > performance issues and its easier to read the TIME FACTOR.. > dont understand why ppl dont get it.. > > ***if it takes 5 seconds to add 1+1, its going to take 10 > seconds to do that TWICE*** > > We can argue about this all day and still not get anywhere..... > What if i had to calculate 50 Miles Radius for 50 concurrent > users? Where do u think the code will run faster? Like i said.. > there is no point try to prove this whole situation.. unless > you are willing to make a fair decision based on FACTS. Facts are slippery things. First of all, in my experience, it's not necessarily true that you'll see proportional differences. That is, if something takes more times to run a million times, that doesn't mean that it'll take more time to run once. Likewise, it's not necessarily true that if it takes five seconds to add two numbers, it'll take ten seconds to do the same thing twice. That's why it's important to run tests "horizontally", with multiple users, rather than "vertically", with a single user iterating over the same task. The two types of tests often give you radically different results; the first type, though, reflects real-world usage and the second doesn't. Unfortunately, it's a lot easier to perform the second type of test, so that's what people do. In any case, though, I'm perfectly willing to admit that CF isn't the ideal calculation engine. Sure, it would be nice if it was better at that sort of stuff. However, many people here seem to be saying, "yes, we need this to be faster", without asking themselves at what cost. Of course, there will be a cost to this, if MM implements it - nothing is free, and for every feature that is worked on, some other feature is neglected. Without infinite resources, it's simply not possible for everything to be perfected. There are lots of things that I think are probably more important, and I'd prefer that they spend their limited resources improving those things. Fortunately, in a well-architected application, you can simply put that sort of logic somewhere else anyway, so I'm not all that upset about it. Dave Watts, CTO, Fig Leaf Software http://www.figleaf.com/ voice: (202) 797-5496 fax: (202) 797-5444 ::::::::::::: dream :: design :: develop ::::::::::::: MXDC 02 :: Join us at this all day conference for designers & developers to learn tips, tricks, best practices and more for the entire Macromedia MX suite. September 28, 2002 :: http://www.mxdc02.com/ (Register today, seats are limited!) :::::::::::::::::::::::::::::::::::::::::::::::::::::: ______________________________________________________________________ Get the mailserver that powers this list at http://www.coolfusion.com FAQ: http://www.thenetprofits.co.uk/coldfusion/faq Archives: http://www.mail-archive.com/cf-talk@houseoffusion.com/ Unsubscribe: http://www.houseoffusion.com/index.cfm?sidebar=lists
