On Tuesday, Oct 8, 2002, at 08:32 US/Pacific, Mark A. Kruger - CFG wrote: > Pardon me for asking but... how does MM handle a situation where > something > worked correctly on CF 5, but now is failing on CFMX? Is that CFMX? > Or do
It could be user code, configuration or CFMX. There have definitely been examples of code that worked-by-accident on CF5 that doesn't work on CFMX - is that a bug? Not really... if the old code was never supposed to work, that was a bug in *CF5* and perhaps the rewrite fixed it. Now, if you built your application on that bug, thinking it was 'valid', then you'll be hurting with CFMX. Examples of configuration (to some degree): - The multiple SQL statements issue is a data source setup issue (use ODBC Socket instead of Microsoft SQL Driver or whatever it's called). It just happened to work with a particular CF5 driver but it wasn't intentional (despite Ben's 'hints & tips'). Change configuration and it solves the problem. Bear in mind also that the JDBC drivers are *third-party*, not Macromedia-developed. - Virtual site setup on IIS. It doesn't work out of the box, you need to run specific configuration to make it work. When you run the Updater, this also seems to be not covered by the standard install and you have to do additional stuff. Is this a bug? I'd say no but some folks will disagree. Are there real CFMX bugs? Of course. Were there real CF5 bugs? Yes. And 4.52 bugs and 4.01 bugs and so on and so on. This is why it's so critical to isolate a repeatable test that exhibits the questionable behavior. I know that's not always easy but it's a fact of life. As Jesse says, if we can't reproduce it, we can't debug it and if we can't debug it, we can't fix it! Case in point: my team has hit the Apache 1.3.26 process death problem. We couldn't reproduce it reliably, even tho' it happened regularly. We spent a lot of time trying to nail it down. No joy. Apache 2.0.39 does not suffer from this problem. Our solution: upgrade to Apache 2.0.39 and implement the two-tier Apache system (documented in my blog). Another case in point: my team hit a CLOB-related problem with the standard CFMX Oracle driver. Again, no reproducible test case. We moved to the 9i Thin Client and hit a different problem. We tried the 9i OCI driver and the problem went away. Now we have to finish rewriting the piece of code that actually relies on the base driver (it's Java, ironically). > I pay to have MM tell me I need to re-write code? See above. If we documented the change and you call it as a bug then, yes, sure you'll get charged (think of it as an "RTFM Tax") but if we didn't document it and you call it as a bug, it's free per Debbie's email. > This "connection reset > by peer" JDBC socket error that's been troubling us (notice I did not > say > "bugging" :) could well be the result of some server setting or SQL > config > option we have overlooked that ODBC/CF 5 ignores, but JDBC looks for. > If > that turns out to be the case, would MM charge for that? I'm with Vern on this - my feeling is that if you have to use one driver over another in order to solve this and WE DIDN'T DOCUMENT THAT, then it's a 'bug' of sorts and it should be free. "I have always wished that my computer would be as easy to use as my telephone. My wish has come true - I no longer know how to use my telephone." -- Bjarne Stroustrup ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~| Archives: http://www.houseoffusion.com/cf_lists/index.cfm?forumid=4 Subscription: http://www.houseoffusion.com/index.cfm?sidebar=lists&body=lists/cf_talk FAQ: http://www.thenetprofits.co.uk/coldfusion/faq Your ad could be here. Monies from ads go to support these lists and provide more resources for the community. http://www.fusionauthority.com/ads.cfm
