"Brian T. Brunner" <brian.t.brunner at gai-tronics.com> wrote: > The compiler changed, it's current behavior is broken from > my point of view! Have you ever used Microsoft Visual Studio? Cygnus (now Red Hat) has 18 years of experience that is a crapload better. > This compiler broken-ness drives a decision to NOT upgrade > to the current (dysfunctional) compiler nor the OS it rode > in on. Again, blame the GCC 2 developers. Sigh, there is a major history lesson that I could interject. Read up on Richard Stallman (founder of GNU, the FSF and original GCC developer), Michael Tiemann (a very early GCC contributor, co-founder of Cy-GNU-s, and currently Red HAT CTO), the "technical specifics" of the GCC releases from 2.7 to 2.8 (and why some distros avoided it), the adoption of EGCS over GCC 2.8, the FSF nod to Cygnus on GCC 3, the GCC 2.91.66 (EGCS 1.1.2) release, the first 2.95 release, and subsequent 2.95.1, 2.95.2, 2.95.3 and, almost counter-productive, 2.95.4 release. And lastly, if you want to see a lot of flamewars -- some of them based on incorrect information / mob opinion (one minor GCC contributor was totally oblivious to some things) -- the GCC 2.96 tag, Red Hat's release of it, and the re-tagging of 2.97 (which became 3.0). Brings up memories from the even earlier GLibC 1, LibC 4, LibC 5 and the return to GLibC 2. There was a lot of incorrect information / mob opinion back then too. Yet we now have a solid LibC with threading, and a very inter-release ABI compatible C/C++ compiler today. ;-> -- Bryan J. Smith | Sent from Yahoo Mail mailto:b.j.smith at ieee.org | (please excuse any http://thebs413.blogspot.com/ | missing headers)