terry
Freshman Member
Posts: 15 
|
Post by terry on Oct 3, 2014 8:43:16 GMT -5
Greetings.
In mainframe ISPF, BNDS 1 1 or BNDS 2 2, etc. is valid, but SPFLite throws the "Illogical BOUNDS operands" error message.
Was this intentional or an oversight?
Thanks.
Terry
|
|
terry
Freshman Member
Posts: 15
|
Post by terry on Oct 4, 2014 9:43:08 GMT -5
I used the single column boundary quite often in ISPF for multiple purposes. I realize that some of the things I used it for could be accomplished by other means (SORT on a single column, FIND x ALL, CHANGE x ALL, EXCLUDE or ONLY, etc.). It just became a personal habit.
|
|
|
|
Post by George on Oct 18, 2014 12:11:17 GMT -5
Robert: Terry:
I've had a look at this change, and simple though it may appear to be, it is not.
Robert and I have been bouncing private messages back and forth, he in favor of doing it, while I have been opposing it.
I'll be brutally honest here, I oppose it solely out of laziness. BOUNDS support is unfortunately one of those items whose code impact is scattered throughout the entire range of Primary and Line commands. Relaxing the BOUNDS parse to allow BOUNDS 1 1 or BOUNDS 10 10 etc. is not the problem. The problem is the review, testing and possible fixup of every other function which BOUNDS support impacts. In many cases this is multiple tests, in both Keyboard Insert/Overtype mode to be performed. This is the epitome of a lot of work for little benefit (1 user?)
There are many other To-Do's waiting which do have more general benefits, I simply can't justify the effort needed here to make this work.
Laziness? Prioritization?
Whatever, your call.
George
|
|
terry
Freshman Member
Posts: 15
|
Post by terry on Oct 19, 2014 14:42:11 GMT -5
George,
I completely understand. I certainly wouldn't try to argue for such a change. Mostly, I was curious why the difference from mainframe ISPF.
Thank you,
Terry
|
|
|
|
Post by George on Oct 20, 2014 9:45:19 GMT -5
Terry: Robert:
As Robert can attest to, ISPF has some unusual quirks (many totally undocumented) in how it handles certain functions (he's re-written some functions totally to 'get it right'). There is an incredible amount of what I call 'kludgy' code in SPFLite to attempt to duplicate ISPF's actions. Cursor positioning after an interaction is particularly difficult, especially when you have multiple line commands and possibly multiple primary commands all being performed, each individual command has it's own idea about where the cursor should go, sorting them all out is maddening sometimes.
Your request was one of those "If we'd done it right in the first place, it's a no-brainer", but we didn't; and back-tracking now is one of those way out of proportion ROI tasks.
George
|
|
