Dear Aurélien, thank you very much for your bug report.
I tried to enable the parallel build in a previous version (3.2.1-1, see changelog), but it failed to build on some plateforms with small amount of RAM.
It generated an "out of memory" error when compiling the medical imaging feature of CamiTK, which is mainly based on ITK (Insight ToolKit) filters, which themselves are intensively based on C++ templates. For instance, on my machine, with -j=9, the compilation needs more than 8Gb of RAM.
Since 3.2.2-1 the parallel build is disabled (see also the debian/rules, where I added some comments in the % rule). All plateforms seem to be able to end the build without an error now, although I completely agree with you: it takes a llllooonnnnggg time.
The only way to fix this would be to enable parallel build only when there is enough RAM.
I am quite a novice, so here are my questions:* Do you know anyway to estimate the amount of memory needed by a compilation? * Do you know if it is possible to check the available memory in d/r and add some kind of test to enable or not parallel build depending on the answer?
Best regards, EP
smime.p7s
Description: S/MIME Cryptographic Signature
