Hi, On Thu, Feb 05, 2009 at 11:40:01AM +0100, Jens Vollinga wrote:
the strange thing is that your original patch did not compile. How can you compile your sources then, anyway?
I've got CLN and GiNaC headers installed in ~/target/x86_64-linux-gnu/include, so '#include <ginac/ginac.h>' happened to work (which is a bad thing).
So, maybe your patch was not complete?
It was complete, but a bit incorrect (specifically, I've messed up #include directives).
Another theory would be that I did something wrong applying the patch. I checked that and it seems unlikely. Most of the patch concerned completely new files, which should cause no problems. The rest of the patch seems fine as well.
I think the patch was correctly applied. Just in a case, I've built GiNaC from from the `official' branch (before sending the previous mail). Still, `make check' shows no errors (or something unusual) at all.
Maybe my "bug fix" that get the includes right introduces this problem?
I doubt it.
Actually, I cannot image this being the case, but if it is, it would mean that the code has some very serious problems itself (and that should be fixed).
annoy users so they don't skip the `make check' step). So I need more info to find out what is going on.
It is hard to give more information, because the code just keeps on running and the data being processed by several involved functions changes continuously.
The information I'm asking for is: 1. Architecture and OS, 2. Compiler version, 3. CLN version, 4. Compiler flags, 5. Any unusual configuration options.
I'll try to pin it down, but I the best thing to do first, I think, is to find out what the difference between your code and the one in the git repository is!
$ git branch -a * master origin/GiNaC-international origin/HEAD origin/experimental_fclasses origin/ginac_1-0 origin/ginac_1-1 origin/ginac_1-2 origin/ginac_1-3 origin/ginac_1-4 origin/master $ git remote -v show origin git://www.ginac.de/ginac.git $ git diff origin/master | wc -l 0 That is, there's no difference. Best regards, Alexei