[Scilab-users] Scilab 6 // Redefinement of permanent variable does not raise an error!?

[Samuel Gougeon]

Hello Dirk,

Le 08/05/2017 à 10:16, Dirk Reusch a écrit :
> Hi Samuel,
>
> Yes, you are right ... I was looking for a message like "error: ....",
> so may be the formatting of the message could be improved?
>
> Anyway, there is another glitch:
>
> ------------------------------------
>
> Startup execution:
>    loading initial environment
>
> --> a=123;predef("all");a=456
>
> Redefining permanent variable.
>
> --> [msg,err]=lasterror()
>   err  =
>
>     999.
>
>   msg  =
>
>   Redefining permanent variable.
>
> ------------------------------------
>
> It seems, that the error number is now 999 and it has
> been or should be 13 (cf. help error_table)?.

As indicated in the Scilab 6.0.0 CHANGES file (section /Feature changes 
and additions/), error numbers are deprecated. I guess that the table of 
error numbers is still distributed just for the record, because Scilab 
and external codes have not been changed to yield proper error messages 
instead of obscure numbers. The deprecation has not been performed in a 
clean way. In the same way, lasterror() should no longer propose 
returning an err output. etc.


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.scilab.org/pipermail/users/attachments/20170508/4bf3433d/attachment.htm>