On Sat, 12 Apr 2008 11:26:26 -0500, Mark Zelden wrote: > > ... > >Of course this is dependent on using the default names of SYS1.SBLS*. >One shop I was at used SYS1.BLS.SBLS* or BLS.SBLS*. I think at one time it >may have been delivered in CBIPO as BLS.SBLS* (like other products) >and some shops just stuck SYS1 in front of the name. I often see >SYS1.ISF.SISF* , SYS1.ISP.SISP*, SYS1.GIM.SGIM* etc. even >though the MLQ is redundant. Hard to change that stuff in a production >environment... well, maybe not hard, just a PITA when there are batch >processes and people with their own clists etc. > A couple years ago, there was a thread here which advanced the opinion that it is harmful for vendors to attempt to influence the customers' choice of data set names for installation of their products. Incredibly, one of the arguments was, "Well, you gotta have standards so programmers know where to find stuff!" You seem to have provided a PITA argument for uniformity.
I believe that ease of installation of "squatty box" products arises from providing customers guidance in this area, and the customers' willingness to follow that guidance. Further benefits for the customer are easy interoperability of products with less configuration labor, and for the vendors greater market success. I do not deny the argument that the customer must be able to choose installation on an alternate path for testing, but only a masochistic customer chooses an idiosyncratic path for production use. Is there somewhere a customer who has an esthetic dislike for "SYS1" and uses an alternative HLQ for production data sets? -- gil ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO Search the archives at http://bama.ua.edu/archives/ibm-main.html
