On 2014-02-21, at 06:17, Peter Relson wrote: > I think everyone agrees that the "no whitespace" rule for continuation of > a non-macro statement is a real pain. > > ... Even if > there is a compatibility argument, some *process option could be added to > say "I want the new rule for my module". > Where would a compatibility argument arise? Perhaos in continuing long strings? But I suspect there's enough wiggle room in the syntax to solve that compatibly, perhaps with an operator for lexical concatenation of string denotations.
But for compatibility with existing MACLIBs, any such *process option should apply exactly to each member in which it appears, not to embedded COPY code nor MACROs. > We've done the following: >> MVC 0(L'QVSIMGLOGICALPARTITIONNAME,R4),* >> QVSIMGLOGICALPARTITIONNAME Move out the LPAR name > I don't understand the syntax. Does the abutting '*' introduce a comment? > I would also question the use of all-upper case. I believe that the > "experts" say that reading mixed case is easier than all-upper, such as > QVSImgLogicalPartitionName. > I feel likewise; apparently the proponents of the Orator typeface feel otherwise. And a radical proposal, easily stated, probably difficult to implement: Allow the significant part of source records to exceed 71 bytes in width. Eliminate much of the need for continuations. -- gil
