Bill Hart wrote: > If you want to use the 4.3 series, then 4.3.0, 4.3.1, 4.3.2 are > definitely extremely buggy (so much so I find them almost unusable). > Another really buggy release was 4.1.2. > > I am also aware of bugs in the 4.4 series. > > Of course these are my own opinions and may not reflect any objective > reality. > > Bill.
Do you have any opinions on what is the least buggy version of gcc? If Sage ever decided to include gcc in the source tree, what version of gcc would be your preference? BTW, rather embarrassingly, it is 4.3.4 that a few people have said is the least buggy. I said 4.3.3, as on the advice of others, that was the binary I got from HP's site for HP-UX. (That was the best they had). I think I will make a build of 4.3.4. I wish the GNU developers would concentrate more on solving bugs, and less on adding extensions. I believe they are developing some framework that will make it easier to create extensions. Somehow, I do not think that is the most important thing they should be worrying about. I've built a Solaris binary of Sage using gcc 4.4.2. Perhaps I'll remake that using gcc 4.3.4. If a compiler bug makes your Word processor mis-behave, you are likely to notice it. The same is not true with maths software. Hence I see a greater need to more seriously consider what gcc versions are used. There might be an argument for Sage to include a set of gcc sources, and build a stable gcc as part of the build process for Sage. Of course, it would add significantly to the size of the download, and also to the time it took to build Sage. Quality does often come at a cost! Dave -- To post to this group, send an email to sage-devel@googlegroups.com To unsubscribe from this group, send an email to sage-devel-unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/sage-devel URL: http://www.sagemath.org
