PDA

View Full Version : A buglet????



David Webber
07-03-2007, 08:51 AM
I have been doing conditional compilations with assorted compiler variables and may have found a small bug.

I have a compiler variable [OFFER_NETWORK_INSTALLATION] and dependent on the release it can be set (with #set compiler variable) to $TRUE$ or $FALSE$

Take a release where it has been set to $TRUE$

If I test it with #ifcompvar for equality with $TRUE$ the test fails. It works fine if I compare it with 1 or with another compiler variable which is set to 1, but not with $TRUE$

Can you reproduce this one - or am I going potty?

Dave

linder
07-03-2007, 09:10 AM
Dave,

Would it be possible for you to send us a small demo project? The constant $TRUE$ should be resolved to "1" during the compilation process. Perhaps this fails.

It would be great if you could send a small demo to support (at) lindersoft (dot) com.

Thank you for your help.

Friedrich

David Webber
07-03-2007, 09:28 AM
Dave,

Would it be possible for you to send us a small demo project? The constant $TRUE$ should be resolved to "1" during the compilation process. Perhaps this fails.

It would be great if you could send a small demo to support (at) lindersoft (dot) com.

Thank you for your help.

Friedrich

I'll see if I can reporoduce it in a small project - the current script is over 500 lines long and installs hundreds of files. [Part of the reason I like SetupBuilder is that it lets me structure this in a pretty clear way!]

Dave

linder
07-03-2007, 10:43 AM
Dave,

It's definitely a compiler problem. Item in review - Ticket [SB#707031]

I'll keep you posted.

Thank you for your help!

Friedrich

linder
07-04-2007, 04:58 AM
Dave,

Fixed, thank you. It was caused by a bug in the #ifcompvar compiler directive.

Thanks for bringing this to our attention!

Friedrich

David Webber
07-04-2007, 06:39 AM
Dave,

Fixed, thank you. It was caused by a bug in the #ifcompvar compiler directive.

Thanks for bringing this to our attention!

Friedrich

Glad to be of help.

Dave