Good points. As to the last - if "Infinity > 0" is compiled, there should be a temp file .as hanging around that we could inspect after the fact - to make sure it: got out to the compiler and looks reasonable.
In my (OSX) system, those temp files got put into: /Users/dda/laszlo/lib/jakarta-tomcat-5.0.30/temp/lzswf9/Users/dda/laszlo/src/svn/openlaszlo/trunk-d/test/build/TestProg The first part of the path (up to /temp/) is the temp directory of my tomcat installation. The part beginning /Users/dda/... represents the full path name of the *.lzx program I am running. Anyway, I typed "Infinity == 0" into the debugger and DebugEvaluate.as contains within it ".... Debug.displayResult(Infinity == 0) ...." . This file is overwritten (and the result recompiled) when I type different expressions in. There is caching, so if I type a previously seen expression again, the file will not be overwritten. Raju, you might hunt for that file and take a look - you'll need to look right after the first evaluation. On Aug 18, 2011, at 8:43 AM, P T Withington wrote: > Curious. > > Is it _only_ "Infinity > 0" that fails? And it fails no matter if you have > typed other expressions into the debugger? And it never fails, no matter > what else you have typed into the debugger, so long as one of the things you > typed was "Infinity < 0"? > > What if you ask "1/0 > 0" instead? > > Note that the SWF debugger has a very simple evaluator that can look up > variables and do '.' and '[]' evaluations locally, but anything more complex > is sent back to the compiler to compile and then loaded and evaluated. It's > possible there is a compiler optimization that is going awry... > > On 2011-08-18, at 07:45, Raju Bitter wrote: > >> Yes, it's repeatable. It's not causing any problems for me, but I was >> just surprised to see the result. >> >> On Thu, Aug 18, 2011 at 12:13 PM, Donald Anderson <[email protected]> >> wrote: >>> Hi Raju, >>> I've never seen this. Is it repeatable within your complex app? >>> - Don >>> On Aug 18, 2011, at 5:57 AM, Raju Bitter wrote: >>> >>> In the SWF10 debugger, when I enter "Infinity > 0", there is no >>> Boolean result value printed out. But when I do "Infinity < 0", the >>> result is "false" (Boolean). Once the debugger has evaluated "Infinity >>> < 0", "Infinity > 0" suddenly resolves to "true". >>> >>> I see that behavior only with the relatively complex app I'm working >>> on, currently running on Trunk (revision 19259). I tested with an >>> empty canvas, and "Infinity > 0" resolves to "true" as expected. >>> >>> Has anyone ever seen a similar behavior with OpenLaszlo? And what >>> might be the problem here? >>> >>> >>> -- >>> Don Anderson >>> Java/C/C++, Berkeley DB, systems consultant >>> >>> voice: 617-306-2057 >>> email: [email protected] >>> www: http://www.ddanderson.com >>> blog: http://libdb.wordpress.com >>> >>> >>> >>> >>> >> > -- Don Anderson Java/C/C++, Berkeley DB, systems consultant voice: 617-306-2057 email: [email protected] www: http://www.ddanderson.com blog: http://libdb.wordpress.com
