Leisner.Henr@Xerox.COM (Marty) (04/22/88)
A number of C compilers (aztec for example) have a reasonable strategy for the float/printf problem. There are two versions of printf -- one in mlib and one in clib. If you're program uses floats, you gotta link in m.lib -- if you never use floats, the c.lib version of printf has judicious ifdefs around the floating point conversions. Seems reasonable. It is unreasonable to have to bring in 30k of floating point support to printf("Error code = %d\n", error); marty ARPA: leisner.henr@xerox.com GV: leisner.henr NS: martin leisner:henr801c:xerox UUCP: nsc!nscimg!amps!marty
decot@hpisod2.HP.COM (Dave Decot) (04/23/88)
> It is unreasonable to have to bring in 30k of floating point support to > printf("Error code = %d\n", error); #if (GRATUITOUS & NITPICK) It is unreasonable for the user of this program to have to bring in 30 lbs. of documentation find out what this particular error code is trying to communicate. :-) :-) :-) #endif Dave "stamp out meaningless error codes; stamp in meaningful messages" Decot hpda!decot
richard@aiva.ed.ac.uk (Richard Tobin) (04/25/88)
In article <13065@brl-adm.ARPA> Leisner.Henr@Xerox.COM (Marty) writes: >It is unreasonable to have to bring in 30k of floating point support to > printf("Error code = %d\n", error); Especially when it would probably take less space to say fputs(error_message[error], stderr); -- Richard -- Richard Tobin, JANET: R.Tobin@uk.ac.ed AI Applications Institute, ARPA: R.Tobin%uk.ac.ed@nss.cs.ucl.ac.uk Edinburgh University. UUCP: ...!ukc!ed.ac.uk!R.Tobin