Path: utzoo!mnetor!tmsoft!torsqnt!news-server.csri.toronto.edu!bonnie.concordia.ca!uunet!wuarchive!sdd.hp.com!elroy.jpl.nasa.gov!decwrl!infopiz!lupine!rfg From: rfg@NCD.COM (Ron Guilmette) Newsgroups: comp.lang.c++ Subject: Re: Let's write a new linker! (Was: Re: Naming Conventions) Keywords: name mangling linking symbol clashes Message-ID: <3946@lupine.NCD.COM> Date: 17 Feb 91 08:29:06 GMT References: <1604@tcs.tcs.com> <474@mole-end.UUCP> <1991Feb7.152041.25151@clear.com> Organization: Network Computing Devices, Inc., Mt. View, CA Lines: 47 In article <1991Feb7.152041.25151@clear.com> rmartin@clear.com (Bob Martin) writes: > >In the future we will likely see more true C++ compilers. These programs >will not have to depend on the C object format. Thus the authors could >generate relocatable object files of a format which specified the types >and classes of objects. >[... text deleted...] >So let me encourage compiler writers everywhere to _think_ about an >alternate format for the object files which could eventually lead us >away from name mangling, and towards a new linker. Two points: First, you note that it would be good if object file formats included some mechanism by which "typing" information about objects could be expressed. Have you ever heard of symbolic debugging information? It generally includes very complete typing information and it goes into the object file. I think it fits the bill pretty nicely. Now what do you want to do with that information? (Perhaps you want a linker that will pay some attention to that information. If so, I'm with you 100%). Second point. You encourage compiler writers to *think* about object files formats. Well let me tell you that they do. Some even do more than that. They join organizations (such as the recently formed UNIX International Programming Languages Special Interest Group) to discuss just this type of issue and to try to evolve commonly used formats (such as ELF & DWARF) to meet the challenges of the 90's (including C++). As a member of the UI/PLSIG myself, I encourage everyone who has an interest in such issues to seek out such groups and to participate in their activities and discussions. P.S. Membership in the UI/PLSIG (and/or its mailing list) is *not* limited to UI members. Membership is open to the public. Contact the chairman, Dan Oldman for additional information. -- // Ron Guilmette - C++ Entomologist // Internet: rfg@ncd.com uucp: ...uunet!lupine!rfg // Motto: If it sticks, force it. If it breaks, it needed replacing anyway.