Robert Osfield wrote: > On Tue, Apr 8, 2008 at 2:31 PM, <[EMAIL PROTECTED] > <mailto:[EMAIL PROTECTED]>> wrote: > > Allright.. I didn't know that was the standard, allways used and > seen ".h" used. :) > > > The problems with standards is that their are jut so many to choose > from... .h is most common for C++ simply from C heritage, but in the > early days of C++ loads of others sprung up in the absence of any clear > definition so .H, .hxx, .hpp and many other variants all turn up in the > wild, there a many of these convoluted variations none of which really > make any sense once you take a step back. When Standard C++ finally > made it out it didn't use any of these convoluted attempts at something > different from C's .h, rather it just dropped the extension entirely. > > Compilers just open files that are specified via #include without making > any assumptions, so you can use absolutely anything you want, you could > use .CPlusPlusHeaderFile if you wished and it'll still compile. Back > in the late nineties I made the choice about extensionless header for > the OSG as it aligns itself with what the Standard C++ headers > convention, rather than going for one of the many > .yetanotherabitaryc++headerextensions that were proliferating at the time. > > Back then I wouldn't have thought that it'd take more than a decade for > IDE's to automatically realise that an extension C++ header file is a > C++ header file.... for all the sophistication of modern software some > really dumb arse things aren't possible... >
Indeed... though the real problem for me is the occasional command line grep... it gets really convoluted to create a find + grep command that will accurately search only C and headers. The directory structure helps a lot there, though. -Paul _______________________________________________ osg-users mailing list [email protected] http://lists.openscenegraph.org/listinfo.cgi/osg-users-openscenegraph.org
