On Sat, Sep 27, 2003 at 04:01:34PM +0200, Jan Nieuwenhuizen wrote: >Corinna Vinschen writes: >>>How does it make it unnecessary? Won't it still cause make to return >>>an error as opposed to actually getting make working? >> >>Yes, but it returns a correct, useful error message. Obviously there >>is a system imposed upper limit of command line length on all systems, >>even if it's 256MB or whatever. So relying on these overlong command >>lines is highly non-portable anyway and at least Cygwin now returns the >>correct message if it comes to that. > >Hmm, maybe you're right and we should fix the installation process, but >this is the first problem we heard of. If it's really that highly >non-portable, then Cygwin is the least obscure UNIX system that doesn't >grok this :-)
There have certainly been other "UNIXes" out there which had small command line length limits. Early AT&T System V releases come to mind. Probably 32K is not a common limit anymore, though. One of the reasons I went to the considerable work of implementing the -X option was to bypass arbitrary limits like this and to bypass the command line parsing that Windows enforces. Binaries mounted with -X pass arguments around in an argv list, just like UNIX. So, while the limits still exist, they should be much larger than what Windows uses. cgf -- Unsubscribe info: http://cygwin.com/ml/#unsubscribe-simple Problem reports: http://cygwin.com/problems.html Documentation: http://cygwin.com/docs.html FAQ: http://cygwin.com/faq/
