[Date Prev][Date Next][Subject Prev][Subject Next][ Date Index][ Subject Index]


Chet: > One could say that after all the boldface are replaced, it
> would almost be safe to replace each remaining MDNM with ;
> however, if we do go into the actual programming, it's possible
> to be more exact.

> XY-> I would prefer to do this with 2 labels
> > XY-> ≪LB1≫ BC se /≪MDBO≫/XC ≪IF≪ER≫≪GL2≫≪EI≫GT BD
[(6X)]  BC se
>  -> /≪MDNM≫/XC {same} ≪GL1≫ ≪LB2> same with ≪MDUL≫

> One of the problems with the above is that (for Xyw 3+ anyway)
> one can go direct from MDBO to MDUL to MDNM, not to mention
> various other MDs that can interrupt. Anyway, going from MDBO
> to MDNM can cut out an intervening mode. An alternative program
> would be as follows:
> > The search (after expanding the display) would be for "≪MD"
> and one would then attach the following two characters to a
> saveget.
> > The main loop falls into two parts: odd and even (odd = the
> first code; even the second).
> > Do a string comparison to determine what you have.
> > If odd and it's an MDNM, either delete it as an unnecessary code
> or stop the program with an appropriate message. If not MDNM,
> replace with appropriate code. Update the counter to "even."
> > Go to search loop.
> > If even and MDNM, replace with the matching code for "odd." If it
> isn't MDNM, either stop the program or insert the appropriate
> matching code from "odd" to its left and treat the present code
> as a new " odd" and continue the search for "even."
> > The ER on the search for "odd" exits the program with a pleasant
> message. One could have any number of additional exits for the
> various "odd" problems.
> Chet,

Hey, I tip my hat to you! Yours is much more "fail safe."