> This technique is especially handy if the existence of one or more of > the non-system LX routines can depend on parameter statements/options > for the software. Lumping it all together in one large System-LX might > not be the best option here.
True enough, unless you get creative with ESR-style vectored linkage on the right hand side of the PC call. Once you've got that knocked, you can add, subtract, multiply and divide one or more discrete PC calls into as many functions as you need. Hence, one LX (and one entry table) does perfectly well for all services and all address spaces. It's what we do in our infrastructure and we never even have to take anything down to add new functionality. Sweet. Now if you need space-switching functionality -into- more than one server space, you're going to need an LX and PC entry table for each target server space. Them's the rules and there's no getting around that, but even then you can use the ESR technique with the PC functions in each server space and that gives you infinite flexibility. Nifty huh? CC ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO Search the archives at http://bama.ua.edu/archives/ibm-main.html
