Date: Wed, 9 Feb 2000 11:19:17 -0800 From: Alfred Perlstein <bright@wintelcom.net> To: "Jonathan H. Ballard" <cybertronix@softcom.net> Cc: Pat Lynch <lynch@bsd.unix.sh>, freebsd-questions@FreeBSD.ORG, tlambert@primenet.com Subject: Re: 3.3 to 3.4 changes in Cplusplus affect FreeBSD source? Message-ID: <20000209111917.W17536@fw.wintelcom.net> In-Reply-To: <38A1966C.21C8199C@softcom.net>; from cybertronix@softcom.net on Wed, Feb 09, 2000 at 08:31:40AM -0800 References: <Pine.BSF.4.05.10002090826230.4543-100000@bytor.rush.net> <38A1966C.21C8199C@softcom.net>
next in thread | previous in thread | raw e-mail | index | archive | help
* Jonathan H. Ballard <cybertronix@softcom.net> [000209 08:51] wrote: > Sound technical advocacy is part of the stand I have taken. The > reasonable factor I hope appears as we chat about C++ vs C and the > affect of FreeBSD source. Could we drop this? There's no 'discussion' to be had, you have 2 choices: a) upgrade the entire system b) install a newer C compiler from the ports tree avoiding a system upgrade. Since for some reason 'a' offends you, then I suggest trying option 'b'. > Pat Lynch wrote: > > > > you need to explain your reasoning a bit more. > > > > Why does the addition of C++ into development of FreeBSD affect security > > and stability? > > > > >From my point of view it appears C++ used to be modularized > and now it is more a tighter knit towards the core > development. C++ is a language with strictness. > This strictness might not be flexible enough stableness. > (more...) say what? > > You are citing tradition rather than sound technical reasoning. Let us > > know the technical reasoning behind it =) > > > > -Pat > > I've already got another email... > > > Alfred Perlstein <bright@wintelcom.net> wrote: > > > How should changes to C++ reflect changes > > > in FreeBSD. To me, from traditions, C++ > > > should be more seperated. > > > > Because you're using the FreeBSD system compiler, we update it along > > with other things. If you don't like this, then try using one of > > the egcs ports in the ports tree. > > > > > Bugzilla reveals this illusionment > > > more... > > > > Urm, the guy is right, we've incorperated fixes into the c++ compiler > > to deal with the problems you are facing, please upgrade to the most > > recent copy of 3-stable. > > > > If you can't do that then merge the fix yourself, it's in our cvs logs. > > > > This belongs on -stable or more likely -questions. > > > I have went out and suggested to many Schools to run FreeBSD. Lets say > those schools cvs'ed FreeBSD 3.3 from the source and got their -STABLE. > These Schools just started to feel 'secure' with their setup. We are > focused to have K-12 feel comfortable and secure with UNIX/FreeBSD. > > Now I know C works. I have used C for a long time now. No problem. > > To me, I do not assimulate C++... > it is not C > > How would someone keep K-12 Schools feel secure > when it has been suggested to upgrade to FreeBSD 3.4 > because Mozilla did not work (or did it?), > and to fix C++. > > (and in the news Yahoo and other sites got DOS attacks... > and they run on FreeBSD...yet just something extra > to advocate) huh? > > More email... > > > Terry Lambert <tlambert@primenet.com> > > You are totally missing the point, with your incomplete log > > message URL. > > > > FreeBSD before the change had a dynamic linker problem that > > affected C++ _and other languages_. I can demonstrate the > > problem in the Java JNDI, FWIW. > > > > This boiled down to a problem in the linking of libraries to > > libraries not enforcing the equivalent of RTLD_NOW. Also, > > FWIW, it is still not completely fixed; there are still ld > > problems that won't show errors until runtime, which would > > take an ld rewrite to fix (this is also apparent in an > > inherent problem in the GNU toolchain, and the coice of the > > EGCS as the compiler technology, when it comes to the per > > thread exception handing stack handling in the libgcc.a; > > Jeremy Allison of SAMBA and I worked on a fix for this, > > and Jeremy submitted the patches, but EGCS had already > > chosen a static rather than dynamic approach to the problem). > > > > Terry Lambert > > terry@lambert.org > > > So do I tell these K-12 Schools that that we have > to 'upgrade' all these systems again, yet they > still not completely fixed. > Or do I tell them not to use C++, Mozilla-M13, etc > until it is fixed. I'd like to see you successfully complain to Microsoft that you can't run Office2000 on ms-dos 2.11. *sheesh* You wan't the latest and greatest software to run and compile? You'll need the most up to date system to do so. > 3.3 to 3.4 does not sound like a reasonable technical upgrade. > It feels more like a downgrade. Why? Are you trolling or is there something fundamentally wrong of your understanding of what the 3-stable branch means? We work very hard to keep 3-stable _stable_ could you explain the obvious deficiencies that occurred between 3.3 and recent 3-stable? We'd like to address these 'issues'. > > Maybe this should be kept to freebsd-advocacy. > No it shouldn't, perhaps freebsd-twilightzone. > (I want to get more comments before I chat more...) I'd rather you just drop the subject, again you have several choices: a) use the port. b) cvsup/upgrade (see: http://www.freebsd.org/handbook/stable.html) c) use the port. d) look at the change that was made: (http://www.freebsd.org/cgi/cvsweb.cgi/src) and try to 'backport' it to your version of 3.3 and recompile it e) USE THE PORT. f) USE THE PORT. g) USE THE PORT. got it? -- -Alfred Perlstein - [bright@wintelcom.net|alfred@freebsd.org] To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-questions" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20000209111917.W17536>