[Jprogramming] When an error is not an error

[bill lam]

I believe this is the intended behaviour and a common practice. You should check for error message immediately after error occurs.


> On 27 Dec 2021, at 11:38 PM, Don Guinn <donguinn at gmail.com> wrote:
> 
> To catch errors in code that I was not aware of due to error handling such
> as try-catch or whatever, I checked the value of the error message
> (13!:12''). But it found the error message is set even though no error was
> signaled. Below is a new J session.
> 
> 13!:12''
> 
> |interface error
> 
> | ferase fs
> 
> |[-127] C:\J903\system\util\boot.ijs
> 
> 
> +'x'
> 
> |domain error
> 
> | +'x'
> 
> 13!:12''
> 
> |domain error
> 
> | +'x'
> 
> 
> erase nl ''
> 
> 1 1 1 1
> 
> 13!:12''
> 
> |domain error: erase
> 
> | erase nl''
> 
> 
> JVERSION
> 
> Engine: j903/j64avx2/windows
> 
> Release-a: commercial/2021-12-16T15:15:09
> 
> Library: 9.03.08
> 
> Qt IDE: 1.9.5/5.15.2(5.15.2)
> 
> Platform: Win 64
> 
> Installer: J903 install
> 
> InstallPath: c:/j903
> 
> Contact: www.jsoftware.com
> 
> I don't see this as a bug that needs fixing but is interesting. I have seen
> this show up in a lab attempting to define a tacit inverse of dyadic
> factorial. The message was set in 13!:12'' but not signaled. But using that
> definition would be an error. Maybe some of these errors should be signaled
> so it could be fixed earlier and more easily.
> 
> 
> And I'm curious as to what that interface non-error is all about.
> 
> 
> Yes, I need to look at my initialization of J to find out why some names
> are defined in a supposedly clear workspace.
> ----------------------------------------------------------------------
> For information about J forums see http://www.jsoftware.com/forums.htm