On 2019-01-30, at 11:02:23, Dave Rivers wrote:
> 
> Regarding the &SYS_HLASM_DATE option, it seems like that doesn’t work
> too well for us (the Dignus assembler) as we different maintenance approaches.
> 
> So this doesn’t help with “portable” code between the two assemblers.
>  
Ouch!  You might just need to fake it.

> I don’t expect that to be a major concern to IBM, but I was wondering if a 
> more
> abstract approach might be better?   Something along the lines of other 
> feature-test
> symbols in other languages?
> 
> That is, when facility X is available, perhaps &SYS_HLASM_FACILITY_X could be 
> defined?
> 
> With this approach, the user has to know that a new facility was available on 
> a particular
> date for example, which PTFs were available at that date.   
> 
> What if the user applies a later PTF and not an earlier one?  And, the 
> facility in question
> is only available with the earlier one?
>  
Hmmm...  I notice that HLASM prints a PTF level in page headers.  This
suggests that information is in a CSECT updated by every PTF.  That,
in turn would require that each PTF either PRE or SUP its predecessor,
so, transitively all its predecessons -- a completely linear sequence
of service.

-- gil

Reply via email to