Sequencing of limits file processing

["M. Lazarus" <mlazarus@xxxxxxxx>]

Barbara, Jon,

 Is the sequencing of limits file processing different on a real S/36 
compared to the 36EE?  My testing shows that when the limits file is keyed 
(in this case LMT14I), the /400 automatically uses keyed sequence, while 
the /36 ignored the access path.  Can you confirm?

 Following is an RPG36 snippet defining the limits file processing:

     FBILDTL  IP  F 512 512L 7AI     2 DISK
     FLMT14I  IR  F  14  14 7        EDISK
     IBILDTL  NS  01
     E    LMT14I  BILDTL

 The file LMT14I consists of 1 to x records.  Each record represents its 
own range of records from the primary file to process (i.e. make visible to 
the program.)

 LMT14I data example:

00222010022299
00111010012099
00333050033305

 By way of application explanation, the BIL# consists of a 5 digit number 
followed by a 2 digit line #.

 So, the first limits record tells the program to print BIL# 222, Line# 1, 
through 222, 99.
The second: BIL# 111, 01 - 120, 99
The third:  BIL# 333, 05 - 333, 05.

 It should (and used to) print in the physical order of the record ranges 
in the limits file.  The logical sequence (if the *limits* file, LMT14I, 
has a key) used to be ignored.  Apparently, in the 36EE, the limits file 
key is being used by the RPG program.

 -mark