On Tue, 27 Mar 2012 06:55:42 -0700, Phil Smith wrote: >Tony Thigpen wrote: >>I have used that method also. But, it has the same problem. Anytime you use >>xedit with either (noprof or with a special, single purpose profile, the >>routines are most likely now broken. It's not something that can be fixed by >>a simple 'just change your profile xedit file'. It's also not something that >>could be addressed by using a system wide profile xedit that calls a userid() >>xedit. > >But my point was that in a macro, you have control over the case; if you stack >stuff, you're making assumptions about case translation. I wouldn't write: > >'command change /this/that/*' > >in a macro if I meant to change THIS to THAT; your technique allows/encourages >such sloppiness if you're assuming CASE U, because stuff will get uppercased >when it falls out of the stack. Yes, the subcommand above will do the same in >CASE U, but it seems clearer to me that it's wrong. > >Maybe this is just me... > What are you following up to? You've posted two replies in this (slightly OT) thread, and I see no antecedents.
But, yes. I advocate a principle of minimal munging. Long ago, I tweaked my PROFILE XEDIT so I could open a binary file, make a one-character change (perhaps in hex) and save it with no collateral damage: No case changes elsewhere in the record; no tabs expanded; no sequence numbers modified. These should be the defaults for minimum astonishment factor. -- gil ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to [email protected] with the message: INFO IBM-MAIN
