1. Bug in call_back when compiled
- Posted by Daniel Kluss <codepilot at gmail.com> Mar 20, 2007
- 509 views
- Last edited Mar 21, 2007
http://www.RapidEuphoria.com/uploads/bug1.zip
2. Re: Bug in call_back when compiled
- Posted by Matt Lewis <matthewwalkerlewis at gmail.com> Mar 20, 2007
- 490 views
- Last edited Mar 21, 2007
Daniel Kluss wrote: > > http://www.RapidEuphoria.com/uploads/bug1.zip For some reason, it's not picking up thing2:thing_hello_rid. I haven't dug, but the c code looks ok, so the obvious suspect is CRoutineId. Matt
3. Re: Bug in call_back when compiled
- Posted by Robert Craig <rds at RapidEuphoria.com> Mar 22, 2007
- 493 views
Matt Lewis wrote: > Daniel Kluss wrote: > > > > <a > > href="http://www.RapidEuphoria.com/uploads/bug1.zip">http://www.RapidEuphoria.com/uploads/bug1.zip</a> > > For some reason, it's not picking up thing2:thing_hello_rid. I haven't dug, > but the c code looks ok, so the obvious suspect is CRoutineId. Yes, I see a bug in CRoutineId(), the run-time routine used by the Translator. From reading the C code, it looks like when the second "thing_hello" is looked up by routine_id() in thing2.e, it would return -1, since it will "see" multiple globals with the name "thing_hello", and no namespace has been specified. However, according to the scope rules, this is not an error. It should match the global in the same file as the call to routine_id(). Multiple globals of the same name, with no namespace qualifier provided, is only an error when they are in other files, not when one is in the same file as the call to routine_id(). The interpreter handles this correctly. I'll fix this tomorrow and check in the change. Regards, Rob Craig Rapid Deployment Software http://www.RapidEuphoria.com
4. Re: Bug in call_back when compiled
- Posted by Robert Craig <rds at RapidEuphoria.com> Mar 22, 2007
- 513 views
- Last edited Mar 23, 2007
Robert Craig wrote: > Yes, I see a bug in CRoutineId(), the run-time routine used by > the Translator. From reading the C code, it looks like when > the second "thing_hello" is looked up by routine_id() in thing2.e, > it would return -1, since it will "see" multiple globals > with the name "thing_hello", and no namespace has been specified. > However, according to the scope rules, this is not an error. > It should match the global in the same file as the call > to routine_id(). Multiple globals of the same name, with no namespace > qualifier provided, is only an error when they are in other files, > not when one is in the same file as the call to routine_id(). > The interpreter handles this correctly. > I'll fix this tomorrow and check in the change. OK, I've checked in a new be_runtime.c. It was an easy fix of just a few lines of C code. Daniel's program works as expected now. Regards, Rob Craig Rapid Deployment Software http://www.RapidEuphoria.com