Greetings -- when trying to build glib2, I've stumbled upon the libintl issue, which I've fixed according to http://trac.macosforge.org/projects/macports/wiki/ProblemHotlist The only speed-up I did was as follows: joined all port names with spaces and ran single command like sudo port uninstall -f atk findutils gawk gettext glib2 gmake gnupg gnutar gnutls gsed libgcrypt neon opencdk subversion subversion- perlbindings texinfo wget whois Then I did port clean --archive <same long line with all port names>, worked fine, too. Then I tried to install them anew with port install -f <same long line> (*) -- and fun started, gettext wanted to get installed again after a few ports, said it can't and halted the sequence. I've just cut off the prefix of the port list and hit enter, and did this until it all had run through. Seemed OK. Unfortunately, I got a problem with iconv when I tried it recently, and I've realized I might have another issue. I got GNAT, a GCC with Ada in it, for Mac from macada,org. It installs into /usr/local/ ada-4.3/, which I symlinked to /usr/local/ada/. I might have left / usr/local/ada/bin/ on the path in the shell where I did the port rebuilds -- I'm no longer sure. The strange thing is, the error I get is /data/bruc/1/1g/unig iconv -f cp1251 -t utf8 < from > to dyld: Library not loaded: /usr/local/ada/lib/libgcc_s.1.0.dylib Referenced from: /usr/local/lib/libiconv.2.dylib Reason: image not found zsh: trace trap iconv -f cp1251 -t utf8 < from > to There's no ibgcc_s.1.0.dylib in /usr/local/ada/lib/ at all! There are /usr/local/ada/lib ls libgcc* libgcc_s.1.dylib libgcc_s.10.4.dylib libgcc_s.10.5.dylib libgcc_s_ppc64.1.dylib libgcc_s_x86_64.1.dylib Questions: -- for the list of ports such as those given by depsearch.sh in ProblemHotlist, how can we see which of them depend on anything from / usr/local/ada/lib/ ? Did it happen due to /usr/local/ada/bin/gcc on the PATH, or somehow else -- and why the non-existent 1.0 dylib? -- should I rebuild those? -- if I rebuild, is it OK to put all names on the line -- then how do I prevent intermediate choking as in (*)? Cheers, Alexy