On Wed, Mar 17, 2010 at 06:08:16PM +0100, Ullrich von Bassewitz wrote: > On Tue, Mar 16, 2010 at 05:56:06PM -0700, Aaron J. Grier wrote: > > I think it would be manageable if the patches were #ifdef'd or > > separated into separate source files. already there is support for > > multiple targets: this would be adding support for sub-targets. > > Please think again. Assume you have a set of replacement files that > use self modifying code. Then you get > > c64 without self modifying code and without ide64 > c64 with self modifying code and without ide64 > c64 without self modifying code and with ide64 > c64 with self modifying code and with ide64 > > If someone requests support for conio using an 80 column driver. This > results in eight libraries: Each of the above with and without support > for 80 columns. And so on ... I do not expect all combinations to be built by default. however, if I configure with "--enable-ide64" and "--enable-selfmodifying" I am asking for the above four combinations to be built. adding "--enable-con80" to the previous two, and I am asking for eight combinations. perhaps there is additional source-level interaction here that I am not understanding? I have been assuming this is a build and link issue, as this is how similar issues are handled with, say, newlib and gcc. let me ask this way: if a patch were to appear which offered per- platform multilib support, would it be accepted for review? -- Aaron J. Grier | "Not your ordinary poofy goof." | agrier@poofygoof.com ---------------------------------------------------------------------- To unsubscribe from the list send mail to majordomo@musoftware.de with the string "unsubscribe cc65" in the body(!) of the mail.Received on Wed Mar 17 19:54:13 2010
This archive was generated by hypermail 2.1.8 : 2010-03-17 19:54:15 CET