Re: CL programming standards

[Peter Dow <petercdow@xxxxxxxxx>]

Use of CRTDUPOBJ would require a variable library name for the TOLIB if 
it's different than the "from" library and not the current library.  In 
some situations, where I want to duplicate an object into the same 
library as another object, I use RTVOBJD to determine what library the 
other object is in, and use that in the CRTDUPOBJ.


On 4/27/2012 4:08 AM, rob@xxxxxxxxx wrote:
> I'll agree that maybe a library name shouldn't be hardcoded.  I have seen
> this in a person's initial program if only to get the name of the data
> area to use which contains their library list.  (Vendor supplied method).
> A variable library name is ok.  Let's say they retrieve the object
> library, file library, etc from this data area.  Now they need to
> copy/reorg/create/etc a file and they want to ensure they get/create it
> from/to the right library.
> Most of the other time we rely upon the library list.  As you've said
> "most", but I say not "all" cases.
> A sending email address like postmaster at myco dot com may be ok but the
> rest we have is table driven.  We don't use snddst to rely upon the system
> distribution directory to be set right.  Too much effort on operations
> when we can merge our Lotus directory into a file on the i to be used for
> email.  And we tie them together with accounting code and something in
> Lotus.  We store employee number in accounting code.
>
>
>
> Rob Berendt