Author: Gian-Carlo Pascutto
Date: 03:12:20 06/19/01
Go up one level in this thread
On June 18, 2001 at 20:14:13, David Rasmussen wrote: >I disagree completely. It rules big time. If people want backwards >compatibility, they can use old stuff. Well if new distributions start shipping with gcc3 and people suddenly can't compile many of the old programs out there you are going to run into trouble. Remember that it is the standards which are changing into incompatible ways. Many times there _was_ nothing wrong with the code itself. Another problem with adhering to new C/C++ standards is that they are breaking away from being interchangable (mostly due to ignorance in the repsective commitees). C99 and C++99 are MUCH more incompatible then C and C++ were. C++ is no longer a superset but a whole different language. Red Hat already started shipping new gcc's and suffered. Sure, what they shipped may have been more standards-compliant than the real stable releases, but people hated it because it couldn't compile old stuff. Hence they chose to include the 'kgcc'. A new major release is indeed the best time to break things if you have to break them. But providing backwards compatibility is even better. And certainly if the old code is only wrong because of a braindead new standard. -- GCP
This page took 0 seconds to execute
Last modified: Thu, 15 Apr 21 08:11:13 -0700
Current Computer Chess Club Forums at Talkchess. This site by Sean Mintz.